To subscribe to our newsletter (no charge) click tom@tomingraminc.com and enter SUBSCRIBE

(or UNSUBSCRIBE to stop)

A black and blue text

Description automatically generated

 

Newsletter

  Home Page   

 

Forgotten Fundamentals

Software Success Lessons (Including AI)

  • Technology Second!  
  • Payback First!  (Results / Cost / Benefit / Risk / Time)
  • 12 Defects, 12 Artifacts, 12 Lessons

 

 

#

Lessons

Summary

999

 

Article in process

 

Lean / Continuous Improvement for B2B Software Product Development

 

A Software Holding Company Based on Danaher,  Berkshire Hathaway, Anduril, Palantir, Amazon Models

 

Symptom:  For 30 Years Software Projects Have Failed to Deliver AS PROMISED 50% to 70% of the time

 

  1.     Fix Root, Root Issues that Prevent Software from Performing As Promised:
    1.     Control and Decisions Vested in COMPETENT People Close to The Work
    2.    Authority / Responsibility Matched
    3.     Scope / Change Management Effective
    4.    Fix The Process First – before Applying Technology

 

  1.     Customer ROI:  Manic About Validated, 10x Value as Defined by Customer.  

 

  1.     Shareholder ROI Yardstick:   Year One Profitability and 20% Sustained Average Return on Invested Capital within Five Years

 

  1.     Incentives and Penalties for Investors, Executives, Team Based On: 
    1.     Present Value Of Future Profits
    2.    Low Invested Capital
    3.     Low Debt
    4.    Long Term Buy and Hold
    5.     Specific Equity Incentives for Those That Build the Business
    6.       NOT based on speculation, hype, “run up stock price and sell out”
    7.    Removing PERVERSE INCENTIVES

 

  1.     Focus & Strategy – Top Five Issues: 
    1.     Sell Only B2B Software Product Development:  Best segment for producing 
      1.       Controlling the ROOT ROOT issues of software:  Control and decisions by right people, authority / responsibility matched, scope / change management effective, fixing the process first – before applying technology
      2.     10x value for customers
      3.  High margins, barriers to competition and low invested capital for owners
      4.   Great quality of work life for software people with fair compensation for value contributed
    2.    Sell Only to Specific B2B Industries Needing Help with Software Product Development
    3.     Sell Only Where Barriers to Competition, Margins and Margin of Safety Will Meet Customer and Shareholder ROI Yardstick
    4.    Compete Only on Best Total Solution (Whole Product – See Sales Execution)
      1.        DO NOT compete on product features, best technology or lowest cost
    5.     Ideal Customer Profile
      1.       B2B Manufacturer, $100 Million to $1 Billion in Sales, real business, real products
      2.     Healthy end customers see need and will pay for high value software products and whole product solutions
      3.  Embraces Lean, Continuous Improvement
      4.  Will acknowledge need for help with software product development
      5.     Investor and executive incentives are based on present value of future profits, long term, buy and hold, happy customers, happy employees
      6.  Generally private / closely controlled companies.  Not public, not private equity controlled.  Not “growth for growth’s sake”.  Not hype the stock and sell out

 

  1.     Obsession With Hands On Execution
    1.     Mission Clarity – Not  Micro Management, Not Making Decisions for People
    2.    Execution is Defined as Systems and Processes! 
    3.     Everything starts with VALUE as measured by CUSTOMER and works backwards
    4.    Focus on Software and Lean Fundamentals that have Consistently Produced Value In the Past  (details below)

 

  1.     Sales Execution: Sold only as a WHOLE PRODUCT Solution to Line of Business Executives (Not IT).  Will include products / solutions for:
    1.     Sales Channel Through Semi-Retired Senior Executives
    2.    Business Expertise / Advisory / Consulting
    3.     Process Reengineering
    4.    Software License
    5.     Integration / Interfaces
    6.       Data Preparation / Migration
    7.    Controlled Customization
    8.    Training
    9.       Cutover / Go Live / Production Support
    10.       Break / Fix, Enhancements, Upgrades

 

  1.     Details of Lean / Continuous Improvement Applied to Software

See www.tomingraminc.com for work samples

 

 

 

 

Additional Detail on Lean / Continuous Improvement Applied to Software

 

REDUCTION OF WASTE for Software:

  •          Making obvious what adds value by reducing everything else. 
  •          Anything customer would not pay for if they knew about it
  •          Motion / Activity Not Adding Value
    •        REWORK from applying technology before FIXING THE PROCESS
    •        REWORK because scope / requirements / change control not defined up front
    •        WAITING for real feedback from real users
    •        SEEKING PERFECTION (100% solution) when a rapid 60% or 80% solution is more effective
  •          DEFECTS (Bugs, missing key features)
  •          OVERBURDEN:  Rework required because of too many or poorly defined User Stories in Agile Sprints 

 

“CYCLE TIME REDUCTION IS EVERYTHING”

 

ACID TESTS To Know When Getting It Right

  •          Right things happen when CEO / execs not in the room
  •          Many experiments, some failures
  •          Validated results from customers
  •          Validated barriers to competition are producing MARGINS that meet ROIC yardstick with MARGIN OF SAFTEY
  •          How CEO / executive time is spent

 

CUSTOMER VALUE TOOLS (more pending)

  •          Voice of Customer
  •          Value Selling
  •          Segmentation

 

PRODUCT LIFE CYCLE MANAGEMENT

  •          Directly connected to ARCHITECTURE

 

GET GOOD AT EXPERIMENTS, TESTING DECISIONS

  •          Top companies are really good at this
    1.     E.g. Jeff Bezos (Amazon) – invested heavily in testing systems
    2.    They make it easy
    3.     They make it high volume
  •          Sets of test customers ready to go
  •          Control groups - to keep tests valid
  •          Failure and invention are inseparable. Live with failures for the upside.

 

DECISIONS

  •          MADE BY TESTING - not executive rank or hunch - whenever possible.
  •          Decision rights are close to the front lines
  •          Decide as late as possible
  •          Leadership DICTATES when necessary, but experiments beforehand to make sure dictating the right things!!!

 

ROOT CAUSE ANALYSIS (Data Based)

  •          Observe to understand problems.
  •          Root cause analysis by data

 

SYMPTOMS – Top Five

  •          Software projects fail to deliver as promised 50% to 70% of the time
  •          Completed projects tend to deliver 60% or less of promised features
  •          CONTINUAL CHANGES result in Scope creep, eat up money, time and patience resulting in UNPROFITABLE projects, stop work and extreme frustration
  •          Delivered software does not match business need and priorities.  Often makes things worse
  •          Bugs, unreliable, unsecure, technical debt, can’t be fixed or upgrade, failed integration…

 

ROOT CAUSES  – Top Five

  •          Work done is NOT THE RIGHT WORK, OUTCOMES NOT RIGOROUSLY SPECIFIED
    •        Requirements / Scope / Change Control Ineffective
    •        Failing to Fix Process First – applying tech to ever changing business process
  •          Value, Payback, Outcomes Not Defined or Delivered as Customer Defines Value
    •        Vendor and Customer failure to define and prioritize outcomes on rational ROI basis
  •          “10x ROI Margin of Safety” Failure:  Absence means CANNOT AFFORD TO DO PROJECT RIGHT.  Margin of safety needed to overcome problems, sustain the effort
  •          Barriers to Competition are not adequate, resulting in price competition that makes effective outcomes and profitable projects impossible
  •          Intangible nature of software and process outcomes makes HUMAN FAILINGS extremely difficult to manage
    •        TOO MANY PEOPLE REWARDED FOR POOR CONTRIBUTIONS

 

ROOT ROOT CAUSES – Top Five

  •          BAD DECISIONS made at WRONG LEVEL by WRONG PEOPLE
  •          FOCUS FAILURES
    •        Focused on TECHNOLOGY when rarely the important issue
    •        Focused on ACTIVITY (because it is comfortable) when OUTCOMES are failing
    •        Focused on “WHAT WE DO” – not Whole Product Solution customer needs
    •        Trying to be TOO MANY THINGS for TOO MANY PEOPLE. 
      •         It is SO HARD TO SELL, most NEVER SAY “NO” to a dollar of revenue
  •          CONTROL, COMPETENCE, AUTHORITY, RESPONSIBILITY not matched with right people at right level
  •          INCENTIVES / PENALTIES NOT ALIGNED.  
    •        Executives not penalized for bad decisions
    •        Unearned Rewards:  Executives, managers, tech people’s resume value increases for just having participated in tech efforts
    •        Vendors, Consultants, Contractors paid with poor connection to outcomes
    •        Short Term Speculative investor / executive mindset more gambling oriented than based on present value of future profits from happy customers
  •          NO LEARNING CURVE ECONOMIES OF SCALE – Always doing things for the FIRST TIME

 

Rapid Prototyping with rapid, direct, yes/no feedback from real users

 

Continuous Improvement Events

  •          Brainstorming
  •          Participation in decisions, experiments, solutions

 

Actively Fight

  •          Unnecessary Complexity, Features
  •          Non-value-add activities, reports, controls
  •          Interference, micromanagement, attempts to control by wrong people

 

Some KEY DEFINITIONS / RULES from Toyota Production System

  •          "Value" is any action or process that a customer would be willing to pay for.
  •          Quality = conformance to specification

 

AGILE KANBAN BOARD – Additional Notes

  •          Make the invisible visible. 
  •          TO DO, DOING, DONE
  •          All work shall be highly specified as to content, sequence, timing, and outcome
  •          Every customer-supplier connection direct with an unambiguous, RAPID way to send requests and receive yes or no responses
  •          The pathway for every product and service must be simple and direct.
  •          Goal:  improvements made by scientific method, guided by a teacher, lowest possible level in the organization.

 

Testing, Quality, Self Inspection Built Into Every Step

 

Technology is Secondary, But Will Be Best Practical Within Above

 

KEY CAUTIONS

  • Assumes strategy is right, producing the right product / service for the right customer with margins and barriers to competition that support yardstick ROIC
  • Risk is being overly focused on “doing things right” instead of “doing the right things”
  • Core weakness of optimized manufacturing is TOO RESISTANT, HARD TO CHANGE , slow to adapt to right strategy changes

 

 

DBS (Danher Business System):  See exhibit 11 HBR’s Danaher Case, Updated 2015  (This is a purchased case.  Tom’s markup available by Zoom call only).

NOTE:  They have a Lean Software Module!

 

 

SUMMARY OF TOTAL RETURNS, LAST 30 YEARS, 9/22/2026 from CLAUDE

  •         S&P 500
  •         BERKSHIRE HATHAWAY
  •         DANAHER 

(Close approximations)

Company/Index

30-Year Annualized Return (CAGR)

30-Year Cumulative Total Return

$10,000 Invested 30 Years Ago Would Be Worth

Danaher (DHR)

~14.9%

~6,397%

~$649,700

Berkshire Hathaway (BRK.A)

~11.2%

~2,250%

~$235,000

S&P 500

~10.3–10.9%

~1,770–1,900%

~$187,000–$200,000

 

 

EXAMPLE OF TARGET CUSTOMERS:  DANAHER B2B INDUSTRIAL SEGMENTS, CUSTOMERS, KEY PRODUCTS

 

See exhibits 1 and 11 for brands, segments and specific customer details  HBR’s Danaher Case, Updated 2015  (This is a purchased case.  Tom’s markup available by Zoom call only).

 

 

  1. Dental, $16 billion market, $1.6 billion in sales
    1. General practitioners
    2. Endodontist
    3. Orthodontists
    4. Lab technicians
  2. Life sciences, $4 billion market, $900 million in sales, operating margins under 10%
    1. Research labs
    2. Hospitals
    3. Universities
    4. HISTOPATHOLOGY Labs
  3. RADIOMETER $12 billion market, $400 million in sales, best operating margins at 25%!
    1. Critical care in hospital
    2. Central Labs
    3. Operating room 
    4. Emergency department
  4. Environmental 
    1. $2.5 billion market, $1 billion sales near 50% market share for Gilbarco Veeder-Root:  Retail point of sale and payment system upgrades petroleum, big box, supermarkets, convenience store, Fuel dispensers, tank gauge systems, merchandising, EV charging
    2. $6.5 billion market, $1.1 billion sales WATER, HACH Chemical, others, water quality, waste water treatment, Automation technology for municipal, industrial, electronics, bio pharmaceutical, boiler cooler, environmental agencies
    3. Test and measurement
    4. $4.5 billion market, $1.3 billion sales FLUKE industrial/electrical, calibration/meteorology, hospitals/field service, communications, data, com,
    5. $12 billion market, $1.1 billion sales, operating margins under 10%  TEKTRONIX. oscilloscopes (invented the oscilloscope), logic analyzers, spectrum analyzers, signal generators, digital multimeters video/protocol test gear.  Computer, consumer, education, government, semiconductor. Wireless, next GEN networks focus
  5. REMEMBER, I NET under Doug Dickerson. Big Telecom / Internet wins due to measurement.
  6. Motion $14 billion market, $1 billion sales Appears spun off in 2016.
    1. 90% sales to OEM‘s.
    2. Targeted industries selling precision, motion, control, servo motors, drives, automation controls for robots, aerospace, and industrial machines
    3. Linear motion systems, ball screws, linear, bearings, actuators
    4. Miniature and micro motors for medical devices, robots, precision instruments
    5. Engine breaking systems for trucks.  Jake brakes
  7. Product identification, $5.1 billion market, $900 million in sales
    1. Printed, other tags to CPG, Pharma, Electronics, automotive, letter and parcel applications
  8. Mechanics hand tools, $4.5 billion market, $900 million sales

 

 

 

Sources: 

  1.       See https://tomingraminc.com for FORGOTTEN FUNDAMENTALS See articles, lessons below:  660 (Danaher), 660.5 (Buffett Lessons) 666 (Amazon) 663 (Architecture Fundamentals) 667 (Practical AI Now: Code Generators, Low Code, No Code Solutions )
  2.  The Surprising Success of Hands-On Leaders, Harvard Business Review, November 2025 by Scott Cook, Cofounder of Intuit and Nitin Nohria, Former Dean, Harvard Business School
  3.  HBR Interview with Scott Cook
  4.  Summary of Lean applied to Software while working for Mitel, being led by a former Danaher executive https://tiainc.net/SkinnY%20Support%20for%20Lean%20Summary.pdf
    1.       NOTE:  This failure was likely due to absence of controlling interest and Danaher executive support for Lean / Continuous Improvement
  5.  Claude:
    1.       Continuous Improvement contrasted with Lean
    2.       30 Year Shareholder Returns for Danaher, Berkshire Hathaway and Fortune 500
  6.  HBR’s Danaher Case, Updated 2015  (This is a purchased case.  Tom’s markup available by Zoom call only).

 

668

 

Article in process

 

Liability as Answer to AI Safety – Rather Than Ineffective Regulation

 

9/28/2026  WSJ, Andy Kessler “The Real AI Fear:  Lawyers”

667

 

Article in process

 

Big Software Success Lessons From Ukraine And Drone Mission Control Room

 

Wall Street Journal article September 26, 2026 inside Ukraine drone mission control room

 

  1. Measuring the RIGHT, FEW things in dashboard on Commander‘s desk.
    1. Russians killed today, EG 343.
    2. Russians killed for each drone command person lost: 656

 

  1. OUTCOMES
    1. Decentralized decision-making:Commander able to delegate 90% of combat decisions
    2. Ancillary, supporting units able to take full advantage of all battlefield information Historical data available for full use
    3. PERVERSE INCENTIVE for falsifying battlefield results/conditions is removed by data from multiple sources
    4. Quick analysis and action Fast weapons improvements through fast, accurate, large scale data / feedback

 

  1. “Spending the smallest amount of budget possible”

 

  1. Battlefield management system, necessity of fighting for their lives and homes, etc. results in trouncing both Russians and NATO allies in exercises

 

  1. Misplaced focus on bright, shiny object:
    1. Western Allies Focused on best drone, which rapidly becomes obsolete RIGHT FOCUS is the “Symphony“ of everything that works together for Ukraine

 

  1. Drones are CHEAP, and GOOD ENOUGH

 

  1. System has the RIGHT USE CASES /VIEWS/capabilities for PERSONAS: commander, mid-level, front line, supporting units, intelligent units

 

  1. Full INTEGRATION of every drone, sensor, shooting platform into single network

 

  1. Current US systems are COMPARTMENTALIZED, ISOLATED, LESS EFFECTIVE against modern threats

 

 

 

 

666

Amazon Lessons

Article in process

 

Big Software Lessons from Amazon

 

From Bruce Allen, Former Dir(?) of Software Training Products

 


Think about PRODUCT versus PROJECT, mindset, operations improvement, projects versus revenue generation product
 

Amazon’s 16 leadership principles from ChatGPT

 

  1. Customer Obsession – Start with the customer and work backwards. Earn and maintain customer trust.
  2. Ownership – Think long term, act on behalf of the entire company, and take responsibility beyond your immediate role.
  3. Invent and Simplify – Innovate while making processes and solutions simpler.
  4. Are Right, A Lot – Exercise strong judgment, seek diverse perspectives, and challenge your own assumptions.
  5. Learn and Be Curious – Continuously learn, improve, and explore new ideas.
  6. Hire and Develop the Best – Raise the talent bar through hiring, mentoring, and developing others.
  7. Insist on the Highest Standards – Maintain exceptionally high quality standards and continually improve them.
  8. Think Big – Create bold visions and pursue innovative opportunities with broad impact.
  9. Bias for Action – Value speed and take calculated risks when decisions are reversible.
  10. Frugality – Accomplish more with fewer resources by being resourceful and inventive.
  11. Earn Trust – Listen carefully, communicate honestly, treat others with respect, and be self-critical.
  12. Dive Deep – Stay connected to the details, verify facts, and investigate problems thoroughly.
  13. Have Backbone; Disagree and Commit – Respectfully challenge decisions when you disagree, then fully support the final decision once it’s made.
  14. Deliver Results – Focus on key priorities and achieve quality outcomes despite obstacles.
  15. Strive to be Earth’s Best Employer – Create a safe, inclusive, productive workplace where employees can grow and succeed.
  16. Success and Scale Bring Broad Responsibility – Recognize that a company’s impact extends beyond its business and act responsibly toward communities, the environment, and future generations. 

 

665

Who Gets The Margin?

Article in process

 

You Need to Invest in AI, but Don’t Expect a Return

 

By Anil K. Gupta, Wall Street Journal 9/1/2026

 

KEY SUMMARY POINTS – EXCELLENT (Mostly)

 

STRATEGY MISTAKE:  INVESTING IN, COUNTING ON ROI WHEN YOU CANNOT CREATE BARRIERS TO COMPETITION

 

  • “Investment in AI should be approached as investment in infrastructure, 

 

  • Paraphrase the Red Queen from Lewis Carroll’s “Through the Looking-Glass,” it may take you all the running to just stay in the same place. 

 

  • Any advantage to be gained will depend on the complementary assets around it. The three that matter most are proprietary data, redesigned workflows and organizational capabilities 

 

  • AI models trained on proprietary data—decades of product formulas, sensor logs from machinery or customer transactions—can produce insights and predictions that off-the-shelf models can’t match. 

 

  • But if the data are equivalent, they yield no advantage. What matters is differentiated, superior data. 

 

  • Using AI to automate individual tasks improves productivity, but these gains are easy for competitors to replicate. 

 

  • Greater advantage comes from redesigning how work gets done across interconnected processes. 

 

  • Such integration is harder to replicate because it requires changes in processes and systems, not merely access to AI models. 

 

  • AI is evolving too rapidly for any model, application or workflow to hold an advantage for long.” 

 

  • “What complementary assets do we possess or can we build that might enable us to capture some of the value if we do?”

664

 

Article in process

 

Lessons from the Bond Market

 

Capital allocation in the real economy became distorted due to artificially low interest rates since the 2008/2009 crisis

 

Created a subsidy for profligate governments.

 

TOM:  AND LUNATIC INVESTING AND ATTITUDES ON TECH / SOFTWARE INVESTMENTS

 

CAUTIONS:  Same effect on entire economy – from colleges expanding their real estate empires to California’s bullet train (which may never be finished).

 

10 year treasury bonds are at 4-5% which is normal.

 

The debt crisis many expected is not here yet – but keep an eye out.

 

Wall Street Journal, 9/2/2026

663

 

Link

Architecture

 

Why No Code / Low Code Solutions Have Limited Effectiveness

Forgotten Fundamentals:  Architecture Checklist

 

Following is a summary of our group discussion from August 24, 2026.  

 

The checklist is intentionally brief – to act as a reminder – not fully discussing each item.

 

Thanks to Cal Smith, a Salesforce Technical Architect, for joining us and helping create this list.

 

Finding Good Work for Senior Software People – Implications

  •           This checklist and the checklist in article 657 “AI Code Generators, Low Code, No Code Solutions” show where Vibe Coding, No Code, User Development and similar approaches run into trouble – and become unviable / impractical. 
  •           This is where we want to look for work – where the magnitude / complexity of the workflow gets beyond AI’s capabilities
  •           NOTE:  It is becoming clear that AI Large Language Models have inherent flaws in producing the same, repeatable output from a set of instructions.  This is a HUGE PROBLEM for Code Generation and Complex Workflows
    •         Preparing your AI Code Generation with a common set of strong instructions is a necessary solution.  “Frameworks” mentioned below are the beginning of managing this need.  More to come…
  •           REGULATED INDUSTRIES may be a place to look for good work because they have less tolerance for “AI slop” and inconsistent outputs that are tolerated by other industries.

 

GENERAL DO’s to Consider

 

  1. Reuse of Code:  Great in theory but very real costs and time required.  Often abandoned in press to deliver

 

  1. TESTING
    1. Unit Test Coverage – built into code packages
    2. Negative Testing – test what should not happen but possibly could
    3. Control of running tests regarding agents
    4. Testing automation 

 

  1. Coding Conventions
    1. e.g. A class can’t be longer than 300 lines
    2. e.g. Class name standards
    3. e.g. Claude- you must tell it what conventions to use (Prompts – reuse and refine)

 

  1. Interfaces
    1. Read only 
    2. Read write
    3. Tools, APIs e.g. Mulesoft making integration much easier

 

  1. Performance under load
    1. Transaction volume

 

  1. End points:  What devices do you have to support?

 

  1. Lifecycle of software:   P.O.C. / Throwaway vs. 10 year+ usage for some applications

 

  1. AGILE Considerations:  e.g. need architecture / solution preparation built into Agile cadence, in advance of the development story sprint.  Usually cannot cram architecture and development into a single two week user story / sprint.

 

  1. Security:  (Huge topic – covered in later sessions) 

 

  1. Technical debt
    1. Old systems – what are we inheriting / must deal with?
    2. New code – what are we creating?  Is it avoidable?

 

  1. Release plan  (Tom’s favorite tool for managing scope / change requests)

 

  1. Upgrade path  (for the application itself plus the underlying tools used)

 

FRAMEWORK EXAMPLES

  1. Error handling framework
  2. Triger class framework
    1. E.G. ...runs after DML type save 
    2. E.G. …rolls back if not completed
  3. Design Pattern / Framework.  Something repeatable, known to work.
    1. E.g. Service Cloud Oriented (Salesforce)
    2. E.g. Microservices
    3. E.g. Design Checklist for developers that Tom created at Elbit with Doug D oversight

 

SOFT ISSUES

  1. REAL LIFE:  Architect CS removed from Covid app project because he objected that solution was not HIPPA compliant
  2. BEST TECHNICAL PEOPLE TEND TO BE INTROVERTS.  After a few bad experiences they become extremely reluctant to speak up
  3. PERVERSE INCENTIVES:  See Tom’s published research.  Too many people benefit while not materially contributing to project outcomes
    1. E.g. Architect ST had multi-million dollar project highjacked by peers advocating pet tool with no substantive benefit

 

GENERAL TOOLS / APPROACH

  1. Object Oriented
  2. Microsoft 
  3. Linux et. al.
  4. Open source
  5. Database
  6. Customize or not
  7. Cloud vs. on premise
  8. Developer environments
  9. No code / low code role

 

DON’Ts / TRAPS TO AVOID

  1. The “FrankenApp” Problem… one person crams too much complexity into a single chunk.  Sometimes due to ego or inexperience.  Sometimes job preservation.  Problem goes by many names:
    1. “God Class” problem
    2. “Rambo Coders” Luis Okhuysen
    3. No Code / Low Code – trying to do too much in a single workflow – Cal Smith and Tom saw examples with Salesforce workflow tools
    4. SOLUTION:  Must break down logically into small understood chunks e.g. psuedocode
  2. A VERY BIG PROBLEM WHEN COST / BENEFIT DOES NOT JUSTIFY DOING IT RIGHT!  Tom saw this devastate a $400 million company
  3. “Go Do” Trap
    1. We will always have time pressure
    2. E.g. Unit testing, integration testing so often under done in rush to production

 

To be continued…

 

662.5

 

Article in process

 

Other Product-First B2B Software Companies Like Anduril

 

From Chat GPT and Claude

 

CHAT GPT

 

What other B2Bsoftware companies are pursuing "develop product first" strategies similar to Anduril? Summarize their target customers and business model

 

Yes. If by “develop product first” you mean the Anduril pattern of building a substantial capability before the customer has written a detailed requirement, demonstrating it in the customer's environment, then expanding from the initial use case, there are several important B2B software companies using variants of it.

I would distinguish this from ordinary product-led growth (PLG). PLG usually means “build a good product, let users try it, then sell more.” The Anduril-style model is more aggressive:

Build a capability → put it in front of a difficult customer → learn from real operations → modify the product → prove economic/operational value → expand the account → turn the solution into a repeatable platform.

 

The closest examples

Company

Primary customers

Product-first characteristic

Business model

Palantir

Governments, defense, intelligence, large enterprises

Builds platforms and deploys engineers into difficult operational environments

Large enterprise contracts + expansion

Scale AI

AI labs, enterprises, governments/defense

Builds AI infrastructure and then develops customer-specific production applications on top of it

Platform + data + application contracts

Datadog

Developers, IT, security, large enterprises

Builds products ahead of individual customer requirements; customers adopt additional products over time

Usage/subscription + land-and-expand

Snowflake

Enterprises, data/AI teams

Built a new cloud data architecture rather than adapting traditional data warehouses

Consumption-based software

MongoDB

Developers and enterprises

Developer adoption precedes enterprise procurement

Subscription/consumption

Atlassian

Developers, engineering, IT, business teams

Product-led adoption rather than traditional enterprise sales

Subscription + cloud + land-and-expand

Samsara

Industrial, transportation, construction, field operations

Develops integrated hardware/software platform around operational problems

Subscription + devices + expansion

CrowdStrike

Enterprises, government, security teams

Product built around a new cloud-native security architecture

Subscription platform + modules

ServiceNow???

Large enterprises

Starts with a workflow product and progressively expands into an enterprise operating platform

Subscription + modules

Scale AI

AI labs, enterprises, governments

Especially close to the Anduril/Palantir model because engineers build directly with customers and then generalize the capability

Platform + services/application revenue

But the degree of similarity varies considerably.

 

1. Scale AI — probably the closest current example

I would put Scale AI near the top of the list if your objective is to identify companies that could follow the Anduril/Palantir pattern.

Scale originally became known for data labeling and AI training infrastructure. It has been moving toward a much broader model: building complete AI systems for enterprises and governments.

Scale now explicitly describes its enterprise business as building, deploying and operating AI systems in production. Its customers include organizations such as BP, Mayo Clinic, Meta, Howard Hughes and government/defense organizations. 

Its particularly interesting tactic is forward-deployed product management.

Scale's job descriptions explicitly say that forward-deployed product managers work inside customer organizations, determine what actually needs to be built, and feed the resulting knowledge back into the core platform. 

Even more revealingly, Scale says:

"What we run on is what we sell."

Its internal AI teams use the platform to solve problems internally and then selectively commercialize the resulting technology. 

Target customers

  • Fortune 500 companies 
  • AI labs 
  • healthcare 
  • energy 
  • insurance 
  • infrastructure 
  • government 
  • defense 

Business model

Platform + application + data + deployment

Scale is trying to move from:

selling AI training data

to:

providing the infrastructure and applications that make AI operational.

That makes it particularly analogous to Anduril's transition from individual products toward Lattice + autonomous systems + Arsenal.

 

2. Palantir — the intellectual ancestor of the model

Palantir is arguably the clearest software example.

Its distinctive strategy is:

Build a powerful general platform → embed engineers with difficult customers → solve real operational problems → generalize the solutions → expand throughout the organization.

Its forward-deployed engineering model is particularly important. Recent reporting on Palantir's FDE organization describes the engineers as an extension of the product organization rather than conventional consultants or customer-success personnel. 

The resulting product architecture is:

Gotham / Foundry

+

Ontology

+

AIP

+

Apollo

The customer is therefore not simply buying an application. It is increasingly adopting a software operating layer.

Target customers

Government

  • defense 
  • intelligence 
  • public agencies 

Commercial

  • manufacturing 
  • healthcare 
  • energy 
  • automotive 
  • finance 
  • logistics 

Business model

Large initial deployments → operational proof → expansion into additional workflows → enterprise-wide platform adoption.

This is the purest software analogue to Anduril.

 

3. Datadog — a much more conventional version of the model

Datadog is interesting because it demonstrates how the product-first → land-and-expand model can work at enormous scale without Palantir's heavy forward-deployment component.

Datadog explicitly describes its model as:

land-and-expand

A customer might initially deploy infrastructure monitoring, then add:

  • application monitoring; 
  • logs; 
  • security; 
  • user monitoring; 
  • cloud security; 
  • network monitoring; 
  • AI monitoring. 

Datadog's investor materials show that customers initially adopt a product and then expand across the integrated platform. 

Target customers

  • software companies 
  • cloud-native businesses 
  • enterprises 
  • developers 
  • IT operations 
  • security teams 

Business model

Consumption/usage + subscriptions + multiple products per customer

This is a less radical version of Anduril:

Build a great product first → get a foothold → develop adjacent products → sell the platform.

The critical difference is that Datadog generally doesn't need to invent a new solution inside each customer.

 

4. Snowflake — product first, then consumption expansion

Snowflake is another strong example.

Instead of adapting traditional enterprise databases to the cloud, Snowflake built a new cloud-native architecture around:

  • elastic computing; 
  • cloud storage; 
  • separation of compute and storage; 
  • consumption pricing; 
  • data sharing. 

Its business model is particularly interesting because customers pay primarily according to actual consumption, rather than buying large quantities of software licenses that might sit unused. Snowflake's investor materials explicitly describe this consumption model. 

Target customers

  • large enterprises 
  • data engineering teams 
  • analytics teams 
  • AI teams 
  • developers 

Business model

Platform → usage → increasing consumption

This creates an unusually strong product-first mechanism:

Get the developers/data team using Snowflake → workload grows → consumption grows → revenue grows.

It doesn't require a traditional “sell $1 million software package” approach.

 

5. MongoDB — developer-first rather than executive-first

MongoDB demonstrates another important variation.

The product is designed to be adopted by developers before the CIO necessarily makes a major procurement decision.

MongoDB Atlas makes this particularly powerful because developers can begin using the managed database with relatively little commitment.

MongoDB reported more than 66,400 Atlas customers as of April 2026, with the company specifically emphasizing on-demand scalability and minimal initial commitment. 

Target customers

  • software developers 
  • startups 
  • technology companies 
  • enterprises 
  • application development teams 

Business model

Developer adoption → production workload → enterprise expansion → consumption

The key idea is:

Don't sell the database to the CIO first. Get the developer to build the next application on it.

That is very different from Anduril's government sales model, but strategically related.

 

6. Atlassian — perhaps the best example of product-led enterprise expansion

Atlassian explicitly says it pioneered product-led growth by:

  • building useful products; 
  • delivering value upfront; 
  • providing frictionless online purchasing; 
  • using transparent pricing; 
  • offering free editions. 

It now has more than 300,000 customers. 

The progression can be:

Jira

→ engineering team

→ entire engineering organization

→ product management

→ IT

→ business teams

→ Confluence / Jira Service Management / other products.

Target customers

Essentially any organization with:

  • software development; 
  • project management; 
  • IT; 
  • collaboration; 
  • knowledge management. 

Business model

Subscription + product-led adoption + cross-sell + enterprise expansion

This is a less "Anduril-like" but highly scalable implementation of the same basic commercial logic.

 

7. CrowdStrike — product architecture creates the expansion opportunity

CrowdStrike is particularly interesting because its strategy resembles Anduril's platform rather than product concept.

It started with endpoint security but built a cloud-native platform on which it could add additional security modules.

The customer can therefore progress from:

endpoint protection

→ identity

→ cloud

→ vulnerability management

→ threat intelligence

→ SIEM/log management

→ other security functions.

Target customers

  • enterprises 
  • government 
  • security operations centers 
  • IT/security departments 

Business model

Subscription platform + modules + expansion

The important strategic principle is:

Make the initial product sufficiently valuable that customers are willing to put more of their security stack onto the same platform.

That is conceptually similar to Anduril's attempt to put more autonomous military capabilities onto Lattice.

 

8. ServiceNow — the "operating system" version

ServiceNow demonstrates what happens when a product-first company successfully expands from one workflow into an enterprise-wide platform.

It originally focused heavily on IT service management.

It then expanded into:

  • HR; 
  • customer service; 
  • security; 
  • operations; 
  • development; 
  • enterprise workflows; 
  • AI. 

McKinsey specifically identifies ServiceNow as an example of a company using land-and-expand and SKU-based development on a common platform. 

Target customers

Large enterprises.

Business model

Start with one workflow → add workflows → expand users → become enterprise platform.

This is probably the best historical example of the "one product becomes an operating system" strategy.

 

9. Samsara — closer to Anduril because software meets the physical world

Samsara deserves special attention.

It combines:

hardware + sensors + connectivity + cloud software + AI

for physical operations.

Customers include:

  • trucking; 
  • construction; 
  • logistics; 
  • utilities; 
  • field services; 
  • manufacturing. 

Its strategy is to start with operational visibility and then expand across multiple workflows.

Samsara's investor materials show that its largest new-customer and expansion transactions increasingly involve multiple products. 

Business model

Connected hardware + recurring software subscription + additional applications

This resembles Anduril much more closely than conventional SaaS because Samsara has to control both the physical edge and the software platform.

The analogy is:

Samsara

Sensors → connectivity → software → AI → physical operations

Anduril

Sensors → Lattice → autonomy → physical systems

 

****10. A particularly interesting new category: AI-native companies

There is now a second generation of companies following the Palantir/Anduril model.

The common pattern is:

Don't sell an AI model. Build the complete workflow that makes AI useful inside a business.

Scale AI is the clearest large example.

But the broader category includes companies building AI-native products for:

  • legal work; 
  • accounting; 
  • insurance; 
  • healthcare; 
  • cybersecurity; 
  • software engineering; 
  • finance; 
  • industrial operations. 

The difference from traditional SaaS is that these companies increasingly start with a workflow and build the technology necessary to automate it, rather than asking customers which software features they want.

This is why forward-deployed engineering is becoming increasingly important. The current literature describes FDE as a product-development methodology, not merely an implementation or consulting function. 

 

CLAUDE

 

What other B2Bsoftware companies are pursuing "develop product first" strategies similar to Anduril? Summarize their target customers and business model

 

Since Anduril's "develop product first" approach — self-funding R&D, then selling a finished, working product rather than being paid to attempt one — is mostly discussed in the defense-tech context, most companies pursuing a comparable strategy are its "neoprime" peers (venture-funded firms selling AI/software-centric systems to militaries). I've grouped them below; note several are software-plus-hardware rather than pure B2B software, since that hybrid is central to how this model works in defense.

 

Palantir Technologies

Target customers: National governments/intelligence and defense agencies (its original base) plus large commercial enterprises in healthcare, manufacturing, and finance.

Business model: Palantir builds and sells software platforms for data integration, analysis, and operational decision-making — Gotham for government/defense, Foundry for commercial enterprise, and AIP for AI orchestration. It monetizes via subscriptions to hosted or on-premises deployments with ongoing operations-and-maintenance services, following an "acquire, expand, scale" pattern where it bears pilot costs upfront and profits once usage scales across a customer's operations.

Shield AI

Target customers: U.S. and allied militaries needing autonomous aircraft that operate without GPS or reliable human control links.

Business model: Shield AI develops the Hivemind autonomy software and platforms like the V-BAT drone, building the aircraft and AI stack itself before contracts are secured, then selling finished systems — its recent $1.5B Series G at a $12.7B valuation reflects investor comparisons to Anduril's trajectory as it wins programs of record like the Collaborative Combat Aircraft.

Helsing

Target customers: NATO-aligned national defense ministries and armed forces in Europe — the company explicitly sells only to democracies.

Business model: Helsing rejected the traditional hardware-first defense model, pursuing a software-first approach, aiming to deliver rapid, upgradeable capabilities across platforms. Its Altra Recce-Strike software platform and HX-2 drone are developed with private capital (starting with a €100 million seed from Daniel Ek's Prima Materia) before being sold or integrated with partner hardware like Rheinmetall vehicles and Saab's Gripen jet, aiming to shift its revenue mix toward more recurring software licensing.

Vannevar Labs

Target customers: U.S. military and intelligence agencies needing foreign-language signals intelligence.

Business model: Vannevar Labs turns foreign communications into real-time intelligence for the military, building its flagship Decrypt product — using AI and large language models to translate, organize, and process intelligence data — before landing large production contracts (a nearly $100 million Defense Innovation Unit award). Notably it reported being profitable on $90 million raised, well before Anduril-scale funding, suggesting a leaner, software-only version of the same self-funded playbook.

Epirus

Business model: Epirus was started by a former Raytheon chief engineer who, rather than wait for a traditional defense contract, raised private cash to spin off development and build prototypes for a new kind of directed-energy "force field" weapon, later converting that self-funded prototype work into an Army contract for counter-drone systems.

Target customers: U.S. Army and allied forces needing counter-unmanned-aircraft-system (counter-UAS) protection.

Saronic

Target customers: U.S. Navy and maritime defense customers.

Business model: Saronic concentrates on autonomous vessels, having secured a $392 million Navy production contract; it's grouped with Anduril, Shield AI and others as part of the "autonomy" cohort of neoprimes drawing large late-stage rounds on the strength of product demonstrations rather than cost-plus development contracts.

Govini and Onebrief (pure mission-software layer)

Target customers: DoD program offices and military planning staffs.

Business model: These are described as leading the "mission-software layer" alongside Anduril and Vannevar Labs — Govini is noted as having the best recurring software economics among defense-tech peers, selling operational/analytics software on a subscription basis rather than hardware, closer to a traditional enterprise SaaS motion but still built and proven before large contracts land.

 

One caveat on scope: with the exception of Palantir, Govini, Onebrief, and (mostly) Vannevar Labs, these are software-defined hardware companies rather than pure B2B software vendors — the "develop first, sell finished product" strategy in defense tech is most visible where autonomy software is bundled with a physical platform (drone, vessel, jammer). If you want, I can search specifically for non-defense B2B software companies using an analogous self-funded, product-first go-to-market (this pattern also shows up in some enterprise infrastructure/dev-tools startups), since the defense-tech framing above is a narrower slice of that broader question.

 

 

662

 

Article in process

 

Palantir Lessons

 

Economist 8/22/2026 Appears significant success reducing admin work and headcount in Britain’s Metropolitan Police and National Healthcare Service (NHS).

 

Police claims “[hundreds of officers on patrol instead of in-office…]”, doing admin work.

 

NHS claims 110,000 extra surgeries performed with same staff (probably exaggerated)

 

Palantir “licenses software and sends ‘forward deployed engineers’ to clients (Jack Prim, Jack Henry former CEO admires this approach.)

 

Notes risk of extreme switching costs (TOM:  Palantir’s upside for a job well done!)

 

Economist thinks it is a “data management company” whose business model predates AI.  They are doing a good job of catching up.

  •         BIG NICHE IS EFFECTIVE INTEGRATION OF GOVT LEGACY SYSTEMS
  •         ANOTHER NICHE:  LAGGARD ENTITIES THAT ARE WAY BEHIND IN DIGITIZATION / MODERNIZATION
  •         ENCOURAGING USER PROGRAMMING ON TOP OF ITS PLATFORMS
  •         “World leading in turning a complex military mission into actionable data flows”

 

40% of $6 bb annual revenues comes from U.S. government, 15% from Britain

 

Peter Thiel is chairman.  Alex Karp is CEO (caution – super salesman criticism)

  •         Overstated capabilities claims

 

 

 

CHAT GPT 9/27/2026

Palantir's business model is more software-centric than Anduril's, but there is a useful parallel: both are trying to change defense procurement by combining proprietary software, close customer engagement, rapid deployment, and unusually aggressive expansion from an initial use case into a broader platform.

The literature on Palantir suggests a business built around four mutually reinforcing platforms—Gotham, Foundry, Apollo and AIP—plus an unusually hands-on sales and implementation model.

1. The basic Palantir business model

Palantir describes its business as providing software that integrates an organization's data, decisions and operations at scale. Its four principal platforms are:

  • Gotham — originally developed for defense and intelligence. 
  • Foundry — enterprise data/operations platform. 
  • Apollo — software deployment and infrastructure-management layer. 
  • AIP — artificial-intelligence platform connecting LLMs and other AI models to organizational data and operational workflows. 

As of 2025, Palantir reported 954 customers and $4.5 billion of revenue, with 54% from government and 46% from commercial customers. 

That is an important change from Palantir's earlier identity as primarily a government/intelligence contractor.

In 2024, the mix was 55% government / 45% commercial, so the company is approaching a roughly balanced government-commercial model. 


2. The central strategic idea: become the operating layer between data and action

The most useful description I've found comes from CB Insights.

Its analysis argues that Palantir is positioning itself as an "orchestration layer" between AI models and enterprise data, rather than attempting to compete directly with companies developing foundation models. 

That distinction is fundamental.

Palantir doesn't need to develop the world's best LLM.

Instead:

OpenAI / Anthropic / other model
↓
Palantir AIP
↓
Palantir Ontology
↓
customer's proprietary data
↓
customer's operational systems
↓
actual decision/action

This potentially puts Palantir in a strategically valuable position regardless of which underlying AI model wins.


3. The Ontology is the key intellectual property

A lot of superficial descriptions of Palantir concentrate on "big data" or AI.

The more sophisticated literature focuses on the Ontology.

Palantir describes the Ontology as the architecture that connects:

  • data; 
  • objects; 
  • relationships; 
  • business logic; 
  • permissions; 
  • decisions; 
  • actions. 

Its current platform documentation explicitly identifies the Ontology as the architectural differentiator of the Palantir platform. 

This creates a powerful business proposition.

A conventional analytics system might tell you:

"Inventory is declining."

Palantir wants the system to understand:

"Inventory at Plant 4 is declining because supplier X is late; production line Y will stop in 36 hours; here are three possible responses; execute option 2."

The distinction is:

data analysis → operational system

That is why Palantir increasingly describes Foundry as an operating system rather than merely a data platform.


4. Foundry is designed to become deeply embedded

This is probably the company's most important commercial tactic.

Palantir does not want to sell a tool that an analyst occasionally opens.

It wants to become infrastructure through which an organization conducts important work.

The 2025 10-K explicitly says Foundry is becoming a central operating system for individual institutions and entire industries. 

The progression can look like:

one problem

→ supply-chain optimization

→ production planning

→ quality control

→ inventory

→ procurement

→ enterprise-wide operations

Once Palantir becomes embedded across those workflows, replacing it becomes considerably more complicated.

This produces what could be called operational switching costs.


5. The "land and expand" strategy

Palantir's sales strategy is unusual.

Rather than trying to sell a giant enterprise license immediately, it ********often begins with a difficult, specific problem.

Then it expands.

The company's own filings say it manages customers at the account level rather than simply by industry, looking for additional long-term opportunities within each customer. 

A simplified version is:

small initial deployment

→ prove operational value

→ additional users

→ additional department

→ additional workflows

→ enterprise deployment

→ additional Palantir products.

This explains why Palantir is so interested in organizations with extremely complicated data environments.


6. Counterintuitively, Palantir deliberately targets difficult projects

This is one of the clearest strategic statements in its SEC filings.

Palantir says its customer-acquisition strategy targets large, difficult-to-execute opportunities where:

  • installation is expensive; 
  • data environments are complicated; 
  • failure is costly; 
  • sales cycles are long. 

Palantir argues that those characteristics actually create barriers to entry for competitors. 

In other words:

Palantir doesn't necessarily avoid the hardest customers. It seeks them out.

That's quite different from a typical SaaS company seeking a simple, repeatable implementation.


7. "Bootcamps" are a major sales innovation

This is probably the most interesting Palantir sales tactic.

Historically, Palantir often spent months embedding teams with prospective customers and conducting pilots.

The company acknowledges that this can require substantial resources and may produce little or no eventual revenue. 

AIP changed that.

Palantir developed AIP Bootcamps, where customers can use their own data and workflows and see working AI applications within days.

Palantir says these bootcamps can produce actual workflows in days rather than requiring lengthy implementation programs. 

This is strategically clever because it addresses the fundamental enterprise-AI sales problem:

Don't explain what AI might do. Demonstrate it using the customer's own data.


8. The "forward-deployed engineer" model

Another distinctive tactic is embedding technically sophisticated Palantir personnel directly with customers.

Rather than:

Salesperson → contract → implementation consultant

Palantir's model historically has been closer to:

Palantir engineer → customer problem → build solution → deploy → iterate.

This makes the sales process partly an engineering process.

It also creates a feedback loop:

customer problem
→ engineer builds solution
→ customer uses it
→ Palantir learns
→ platform improves
→ solution becomes reusable elsewhere.

This helps explain why Palantir has historically been willing to spend unusually heavily on individual customer engagements.


9. The company is moving from bespoke deployments toward repeatable platforms

This is one of the most important changes in the business model.

Early Palantir could look somewhat like a high-end technology consultancy.

The strategic objective now is much closer to:

build once → configure repeatedly → distribute widely.

Palantir's 2025 filing explicitly discusses industry operating systems, with examples including:

  • airlines; 
  • healthcare; 
  • automotive; 
  • telecommunications; 
  • shipbuilding; 
  • insurance; 
  • space; 
  • security/risk management. 

The idea is to develop solutions that can be distributed across an industry rather than rebuilt independently for every customer. 

This could materially improve scalability.


10. AIP changes the economic opportunity

AIP gives Palantir a way to monetize the enormous interest in generative AI without having to develop its own foundation model.

The company's architecture allows organizations to use different LLMs while connecting them to Palantir's data, permissions and operational workflows. 

That produces a potentially attractive strategic position:

Model layer

OpenAI
Anthropic
Google
open-source models
customer-developed models

↓

Palantir

AIP + Ontology + Foundry/Gotham + Apollo

↓

Enterprise

Data + workflows + decisions + actions

Palantir therefore has an incentive to remain relatively model agnostic.


11. Apollo is more strategically important than it initially appears

Apollo is essentially Palantir's deployment infrastructure.

It allows Palantir software to operate:

  • in commercial cloud; 
  • government cloud; 
  • on-premises; 
  • classified environments; 
  • disconnected environments; 
  • highly constrained/rugged environments. 

Palantir describes Apollo as a control layer that continuously manages software updates, security and configurations. 

This creates an interesting three-layer architecture:

Layer

Product

Function

Data/operations

Foundry

Enterprise

Intelligence/defense

Gotham

Government/defense

AI

AIP

AI + decisions

Deployment

Apollo

Run/update everything

That architecture is part of what makes Palantir different from an ordinary analytics vendor.


12. Palantir's government strategy is also a commercial strategy

Palantir's government origins are not simply historical baggage.

They have become a competitive advantage.

The company developed its technology in extremely difficult environments:

  • intelligence; 
  • counterterrorism; 
  • military operations; 
  • classified networks; 
  • highly restricted data. 

That gives Palantir credibility when selling to large commercial organizations with similarly difficult data problems.

The company's own filings say it began with U.S. intelligence and counterterrorism and subsequently found similar data problems in commercial enterprises. 

This produces a useful flywheel:

Government

→ difficult problems

→ sophisticated software

→ operational credibility

→ commercial applications

→ additional revenue

→ more R&D

→ better government capabilities.


13. Palantir is deliberately building a dual-market company

This distinguishes it from many defense-tech companies.

In 2025:

Government: 54%
Commercial: 46%

of revenue. 

The company therefore has two enormous markets available to it.

Government

  • defense; 
  • intelligence; 
  • law enforcement; 
  • public health; 
  • disaster response. 

Commercial

  • manufacturing; 
  • automotive; 
  • energy; 
  • pharmaceuticals; 
  • healthcare; 
  • supply chain; 
  • finance; 
  • aerospace. 

The same underlying architecture can serve both.


14. Cloud partnerships are another distribution tactic

Palantir has increasingly partnered with major cloud providers.

The 10-K explains the logic directly: cloud companies already have relationships with essentially every major enterprise and large sales organizations, providing Palantir with an additional distribution channel. 

That changes the sales equation from:

Palantir must find every enterprise customer itself.

to:

Palantir can use cloud-provider relationships to reach enterprises already using those clouds.


15. Developer Tier is a funnel strategy

Palantir has also introduced Developer Tier, allowing developers to experiment with Foundry and AIP with limited upfront commitment.

The company's stated objective is to allow developers to explore and build without significant enterprise costs. 

This resembles a classic software strategy:

free/low-cost experimentation

→ developer adoption

→ proof of concept

→ departmental deployment

→ enterprise contract.

That is quite different from Palantir's historical sales model and could make the company substantially more scalable.


16. Palantir's major strategic advantage: customer data + ontology + workflows

The competitive moat is arguably not the AI itself.

It is the combination of:

customer data

  •  

Ontology

  •  

business logic

  •  

operational workflows

  •  

security/access controls

  •  

AI

  •  

deployment infrastructure

Once those pieces are integrated into a customer's operations, replacing Palantir means replacing much more than a database.

This is why Palantir's strategy is better understood as enterprise operating infrastructure than simply "AI software."


17. The main criticism: it can be expensive and labor-intensive

There is an important contradiction in Palantir's model.

The company wants:

high-margin scalable software.

But its historical sales model requires:

highly skilled engineers working closely with individual customers.

Palantir itself acknowledges that customer acquisition and deployment costs occur before revenue is recognized and that pilots can generate little or no subsequent revenue. 

The strategic challenge is therefore:

Can Palantir convert an unusually labor-intensive implementation model into a genuinely scalable software business?

AIP Bootcamps and Developer Tier appear to be attempts to solve exactly that problem.


18. Another risk: concentration

Although Palantir now has hundreds of customers, the largest customers are still economically important.

In 2025 the average revenue from its 20 largest customers was approximately $93.9 million, compared with $64.6 million in 2024. 

That is both an advantage and a risk:

Advantage: enormous expansion potential within existing accounts.

Risk: substantial dependence on a relatively small number of very large organizations.


19. Government procurement is a structural risk

Palantir explicitly warns that government procurement policies could change in ways that affect:

  • pricing; 
  • contract structure; 
  • margins; 
  • competition; 
  • procurement criteria. 

It also identifies bid protests and budget constraints as risks. 

This matters because Palantir's government business remains more than half of total revenue.


20. Political and sovereignty concerns

There is also a substantial body of criticism concerning Palantir's role in:

  • immigration enforcement; 
  • policing; 
  • military operations; 
  • intelligence; 
  • national health-data infrastructure. 

Recent UK reporting illustrates the issue. Five UK police forces ended a Palantir pilot after two years, citing funding limitations and insufficiently defined benefits; the decision also occurred amid broader concerns about procurement and accountability. 

The criticism is strategically relevant because Palantir's business model depends on deep integration into government institutions.

The same integration that creates a competitive moat can create political and regulatory opposition.



22. Publications I would prioritize

If you're studying Palantir as a business strategy case, rather than simply following its stock or political controversies, I'd put these at the top:

1. Palantir's 2025 10-K

The most important primary source. It lays out the business model, customer acquisition strategy, bootcamps, direct sales force, industry operating systems, cloud partnerships and government/commercial mix. 

Palantir 2025 SEC filing

2. CB Insights — From the battlefield to the factory, Palantir is going on an AI offensive

Useful independent strategic analysis of Palantir's investments, partnerships, AI strategy and movement toward an orchestration layer. 

CB Insights analysis

3. Palantir's 2024 10-K

Particularly useful for seeing how the AIP Bootcamp and Developer Tier strategy emerged and how Palantir's commercial strategy evolved. 

Palantir 2024 SEC filing

4. Palantir's platform documentation

Useful for understanding the actual architecture rather than the investor presentation: Foundry + Ontology + AIP + Apollo. 

Palantir Platform Overview

5. Recent Financial Times reporting

Useful for understanding the political, government-contracting and international risks surrounding Palantir's expanding role in government systems. 


Bottom line

The publications suggest that Palantir's real strategy is considerably broader than "selling AI software."

Its model can be reduced to this sequence:

1. Enter a difficult, high-value problem
↓
2. Embed engineers with the customer
↓
3. Integrate fragmented data through the Ontology
↓
4. Put the resulting workflow into Foundry/Gotham
↓
5. Add AIP and AI agents
↓
6. Connect those decisions to actual operations
↓
7. Deploy and continuously update through Apollo
↓
8. Expand from one use case to the entire organization
↓
9. Generalize the solution into an industry operating system

That explains why Palantir's most important strategic asset may not be its AI models at all. Its objective appears to be controlling the layer where an organization's data, AI models, decisions and operational actions meet. CB Insights' description of Palantir as an orchestration layer between AI models and enterprise data captures this particularly well. 

And that is also where the comparison with Anduril becomes particularly interesting: both companies are pursuing a platform strategy, but Palantir is trying to own the digital operating layer while Anduril is trying to own both the digital layer and the physical autonomous systems beneath it.

 

CLAUDE 9/27/2026

 

A cluster of business-analysis publications describe Palantir's model consistently, with a common framework and some differing emphases.

Core framework: "Acquire, Expand, Scale"

The most repeated structural description, appearing in FourWeekMBA, The Strategy Story, and Business Strategy Hub, is a three-phase model: the company bears the pilot costs in the acquire and expand phases, and it runs at a loss, while in the scale phase, the customers' contribution margins become positive. FourWeekMBA frames the acquisition phase as deliberately targeting difficult customers: Palantir's acquisition strategy targets large-scale, hard-to-execute opportunities at large government and commercial institutions — the kind of customers that most enterprise companies might avoid, due to the high acquisition cost, and difficult technical feasibility — and this acquisition strategy, in theory, works also as an entry-barrier to potential competitors. The Strategy Story adds that in the scale phase, Palantir's investment costs relative to revenue generally decrease, while its software's value increases as the platform's usage increase across the customer's operations. Palantir Business Model Explained (2026) - FourWeekMBA +2

Two (now three) platforms, and how they monetize

Sources agree on the product core: Palantir has three principal software platforms: Gotham, Foundry, and Apollo, providing critical infrastructure to integrate customers' data and operations and run their software in virtually any environment, with the vertically integrated nature of Gotham and Foundry allowing users of varying technical abilities to collaborate — from data engineers to senior decision-makers. Monetization runs through subscriptions to access its software in a hosted environment with ongoing O&M services ("Palantir Cloud"), and software subscriptions in customers' own environments with ongoing O&M ("On-Premises Software"). PitchGrade's more recent research note frames this as a long-term platform licensing business augmented by professional services, historically delivered by "forward deployed engineers" (FDEs) embedded directly in customer operations to build workflows and drive adoption — a model PitchGrade calls capital-intensive and limiting to scalability, which is why the transition to AIP bootcamps and standardized product onboarding is strategically critical. How does Palantir make money: Business Model & Competitor Analysis +4

The AI pivot (AIP)

The Strategy Story's 2026 pieces emphasize that Palantir is positioning itself not as a foundation-model developer but as connective infrastructure: Palantir's opportunity is not based primarily on developing a single foundation model — instead, it seeks to provide the infrastructure through which different AI models can safely interact with enterprise data and real-world operations, addressing the enterprise need for controls over which data models can access, what users are allowed to see, what actions can be taken and how AI outputs connect to existing systems. The same source describes a sector-templating tactic: by developing sector-level architectures, Palantir can potentially accelerate deployment and increase the number of customers served without recreating every solution from the beginning — a strategy that can also create network effects at the ecosystem level. Palantir Business Strategy 2026 - The Strategy Story +2

Revenue mix and growth

Figures cited across sources show the government/commercial split shifting over time. In 2023, Palantir's revenue was $2.23 billion, with commercial revenue growing 20% year-over-year to $1.0 billion and government revenue growing 14% year-over-year to $1.2 billion. By 2025, per The Strategy Story's later piece, government customers generated $2.40 billion of Palantir's $4.48 billion of 2025 revenue, and the company's basic mechanism is described as: customers pay to access and use its software platforms, either in Palantir-hosted environments or within their own infrastructure, generally entering contracts ranging from one to five years, with revenue recognized over the contract term. How does Palantir make money: Business Model & Competitor Analysis +2

Ecosystem partnerships and distribution

Newer strategic commentary (Umbrex, The Strategy Story 2026) highlights partnership-based distribution as a growing tactic: Palantir has publicly expanded its alliances with major infrastructure and enterprise software partners, including cloud providers, which strategically helps distribution, deployment credibility, and access to customers that may prefer a broader ecosystem approach rather than a standalone vendor relationship. The Strategy Story frames this as combining Palantir's software with the market access, industry relationships and delivery capabilities of established partners, while stressing that a defining element of Palantir's business strategy is its emphasis on the lifetime potential of a customer relationship. Palantir Strategy and Business Model +2

Culture and long-game framing

Older strategic profiles (Think Insights) emphasize company culture as a competitive input, describing a strong mindset derived from its founders around expanding and improving services — not only software and machines but employee capacities — combined with an expansion strategy built on constant development through new technology or buyouts of minor enterprises, and a strong emphasis on discretion and secrecy. The Strategy Story ties Palantir's underlying growth thesis to a belief about institutional failure: Palantir believes that the systemic failures of government institutions to provide for the public — fractured healthcare systems, erosions of data privacy, strained criminal justice systems, and outmoded ways of fighting wars — will continue to require both the public and private sectors to transform themselves, thus increasing the need for its products. Think InsightsThe Strategy Story


Sources: Umbrex, FourWeekMBA, Think Insights, The Strategy Story (three separate 2024–2026 pieces), Business Strategy Hub, seat11a.com, and PitchGrade Research.

 

661

 

Article in process

 

Anduril Lessons

 

Watch the Defense Industry startups – especially in Silicon Valley.  These startups have found a way to DRAMATICALLY SHORTEN THE GOVERNMENT PROCUREMENT PROCESS.  

 

Major defense contractors sell programming / software / product development programs to the government – a long, arduous, problem prone process.  

 

Startups use AI to rapidly create complete products and offer them to the U.S. government.  The government says “yes or no”, DRAMATICALLY SHORTENING PROCUREMENT TIMES! 

 

Claude 9/27/2026

 

Core model: self-funded R&D, fixed-price products

The most repeated theme across sources is that Anduril inverts the traditional defense-contracting model. Rather than waiting for a government Request for Proposal and getting paid "cost-plus" for effort (the traditional model, where contracts can only generate 7-12% margins, incentivizing contractors to make projects as expensive and long as possible), Anduril takes on the R&D burden upfront, offering pre-developed, cutting-edge solutions to the DoD and allied militaries, which lets it reach market much faster than the multi-year government acquisition process. Think Insights frames this starkly: most defense companies get paid to attempt a solution, not to deliver one, whereas Anduril spends its own money on R&D first, then sells a finished, working system at an agreed price — absorbing the cost if a system fails, and keeping efficiency gains as margin rather than having them negotiated away in follow-on contracts. Report: Anduril Industries Business Breakdown & Founding Story | Contrary Research +2

 

Lattice as the software moat

Multiple sources identify the Lattice software platform as the real economic engine. Business Model Analyst argues Lattice is an AI platform that ties every product into one command-and-control layer, creating recurring, high-margin, sticky revenue — hardware like drones, towers and submarines are really just delivery vehicles, with the software as ***** the moat and margin. Contrary Research quotes CEO Brian Schimpf's own framing: he's described it as "an à la carte umbrella where everything works together, very similar to AWS," with different components available as needed, and notes the platform's use extends into civilian industries too. Business Model AnalystContrary Research

FedSavvy Strategies places this in the "Neo Prime" category — a new breed of Silicon-Valley-born defense contractor that prioritizes a "product-first," software-centric mindset over the traditional "exquisite hardware" model, investing heavily in private R&D on commercially available products like Lattice, and argues Anduril is evolving from a disruptive startup into a mainstream, established defense prime, with Lattice as the "glue layer" moat and its "Arsenal-1" manufacturing vision providing industrial mass for sustained conflict. FedSavvy StrategiesFedSavvy Strategies

 

Acquisitions as a build-out tactic

Think Insights and others describe a consistent acquisition tactic: since 2021 Anduril has acquired capabilities in small drone technology, autonomous undersea vehicles, solid rocket motors, radar, ruggedized computing and infrared sensors, folding each purchase into the Lattice ecosystem rather than running it as a separate business line, with subsidiaries like Area-I, Dive Technologies, Adranos and Blue Force Technologies supplying drone, undersea, propulsion and aircraft technology that feeds directly into Lattice. A Medium deep-dive and Substack piece both note the Microsoft IVAS AR-goggle takeover as part of this pattern. Think InsightsThink Insights

 

Capital structure

Sources agree venture capital is essential to sustaining the model. Venture investors led by Thrive Capital, Andreessen Horowitz and Founders Fund fund the research and factory buildout that government contracts alone do not cover, while the cap table also includes General Catalyst and Lux Capital. On IPO plans, Anduril has said it isn't planning to go public yet, with Palmer Luckey stating he wants public markets to better understand the business model first — possibly in 2026 or later. Anduril's Fixed-Price Defense Model | Think Insights +2

 

Skeptical counter-take

Not all coverage is admiring. War on the Rocks' "Cogs of War" newsletter is more critical, arguing that for Anduril and peers like Helsing, Saronic, and Shield AI, the real business model may be to get the end-user hooked on cheap hardware, sold for minimal or no profit, so that the company's software later becomes indispensable — generating revenue through permanent licensing and updates. It also flags a structural risk: even though these companies' software suites should theoretically interoperate under the Pentagon's "modular open systems approach," a resulting "zoo of systems" is neither scalable nor especially profitable. War on the RocksWar on the Rocks

 

Financial risk

Business Model Analyst is most pointed on the risk side, calling it a "high-wire act": a $61 billion valuation sits on a company posting a billion-dollar-plus operating loss, betting everything on a gigafactory never built at this scale, serving essentially one customer whose priorities can shift overnight — and the same vertical integration and self-funding that produce the margins also concentrate the risk. Business Model Analyst

 

Sources: Sacra, Contrary Research, War on the Rocks ("Cogs of War"), Think Insights, a Medium deep-dive (BUVCG Research), FedSavvy Strategies (two posts), Business Model Analyst, and a Rigatoni Capital Substack post.

is anduril public

 

CHAT GPT 9/27/2026

 

There is a fairly substantial body of writing on Anduril, but it falls into two different categories: descriptions of what Anduril says it is doing and independent analyses of why the model might work—or fail.

The publications collectively portray Anduril as trying to build something closer to a Silicon Valley technology platform plus industrial manufacturer than a conventional defense contractor.

1. The core Anduril business model

The most consistent theme is that Anduril attempts to invert the traditional defense-contractor model.

Traditional model

Large defense primes generally:

  • receive government requirements; 
  • bid on programs; 
  • obtain government-funded development; 
  • build highly customized systems; 
  • operate through extensive subcontractor networks; 
  • frequently use cost-reimbursement or cost-plus arrangements. 

Anduril's model

Anduril instead:

  1. Identifies a military problem before the government has necessarily written a formal requirement. 
  2. Uses its own venture capital to develop a working product. 
  3. Demonstrates an actual system rather than primarily presenting a proposal. 
  4. Gets an initial military customer.
  5. Iterates rapidly based on operational feedback.
  6. Sells the system on relatively predictable/fixed-price terms. 
  7. Uses the resulting contracts and investor capital to develop the next generation of products.

The Los Angeles Times describes this explicitly as a Silicon Valley approach: Anduril uses software, venture capital and private development capital to build products before asking the government to pay for development. 

Palmer Luckey has similarly described Anduril as a "defense-product company" that puts its own money into building things and arrives with a working product rather than asking taxpayers to bear development risk. 

A 2023 strategy analysis calls this the fixed-price defense model: privately fund development, deliver a finished product, and charge a predetermined price rather than billing for the government's development costs. 

2. Lattice is arguably the most important strategic piece

A number of publications make an important distinction:

Anduril is not fundamentally a drone company.

It is increasingly a software/platform company that happens to manufacture autonomous weapons and sensors.

Anduril describes Lattice as the software platform underlying its "software-defined weapons," integrating Anduril and third-party sensors, vehicles and effectors. 

The strategic significance is substantial.

Instead of selling:

Drone A + radar B + interceptor C

Anduril wants the customer to buy into:

Lattice + an ecosystem of autonomous systems

That gives Anduril an opportunity to become the integration layer connecting numerous military systems.

CB Insights makes essentially this argument, describing Lattice as a digital backbone that can connect systems across land, sea, air and space. 

An academic analysis published in Science as Culture goes further, arguing that Lattice could become an infrastructure through which future procurement decisions are organized. In that interpretation, once a military organization adopts Lattice, additional Anduril products become easier to integrate, potentially creating ecosystem effects and switching costs. 

That is an important strategic distinction from simply selling hardware.


3. "Hardware-enabled, software-defined"

The underlying architecture can be summarized as:

Lattice
↓
common software / data / command architecture
↓
many autonomous platforms
↓
common manufacturing architecture

That allows Anduril to reuse software, autonomy algorithms, communications, interfaces and manufacturing processes across different products.

For example, the same general architecture can support:

  • counter-drone systems; 
  • autonomous aircraft; 
  • loitering munitions; 
  • underwater vehicles; 
  • surveillance systems; 
  • air-defense interceptors; 
  • command-and-control systems. 

Anduril's own description emphasizes that Lattice is sensor-, network- and system-agnostic. 

This is analogous to a technology company's strategy of building a common operating system and then putting many products on top of it.


4. Arsenal: the other half of the strategy

This is probably the most significant evolution in Anduril's business model.

Originally, Anduril's advantage was largely:

software + AI + rapid engineering

It is now attempting to add:

software + AI + vertically integrated mass manufacturing.

Anduril calls the manufacturing architecture Arsenal.

Its announced Arsenal-1 facility in Ohio is intended to become a very large, software-controlled manufacturing operation. Defense News reports that Anduril's objective is "hyperscale" production—tens of thousands of autonomous systems rather than the relatively small production runs characteristic of many sophisticated defense programs. 

The strategy has several components:

A. Design for manufacturing from day one

Instead of designing an extraordinarily sophisticated weapon and figuring out how to manufacture it afterward, Anduril wants engineers to design the product around:

  • simplicity; 
  • modularity; 
  • readily available components; 
  • ease of assembly; 
  • ease of repair; 
  • scalability. 

Janes' reporting on Arsenal-1 describes Anduril designing products from the beginning to be "easily manufacturable, easily scalable, and easily fixable." 

B. Commercial components

Anduril says roughly 90% of its products can use commercially available components and materials.

The objective is to avoid defense-specific supply chains wherever possible. 

C. Flexible factories

Rather than constructing a factory dedicated to one weapon, Anduril wants production lines that can be reconfigured.

Janes reports that Anduril deliberately minimizes fixed infrastructure so that production lines can be expanded, reduced or reconfigured for another product. 

D. Software-controlled manufacturing

Arsenal is intended to connect:

  • product design; 
  • engineering; 
  • bills of materials; 
  • supply chain; 
  • manufacturing; 
  • testing; 
  • inventory; 
  • production planning. 

Anduril describes this as a software-defined manufacturing platform. 

This is important because it potentially makes the factory itself a technology product.


5. The strategic objective: affordable mass

Anduril's executives repeatedly emphasize mass rather than extremely expensive "exquisite" platforms.

The logic is straightforward:

A military cannot necessarily afford to lose hundreds of $100-million aircraft.

It potentially can afford to lose large numbers of relatively inexpensive autonomous systems.

This is particularly important because Ukraine has demonstrated the importance of:

  • large numbers of drones; 
  • rapid replacement; 
  • electronic warfare; 
  • autonomous systems; 
  • inexpensive sensors; 
  • rapid software iteration. 

A European Journal of International Security analysis quotes Anduril's own strategic presentation describing the objective as "affordable distributed mass." 

So Anduril is attacking two traditional defense assumptions simultaneously:

Old model

Make a small number of extraordinarily capable systems.

Anduril model

Make large numbers of sufficiently capable autonomous systems and continuously improve them through software.


6. Rapid iteration is a deliberate tactic

One of Anduril's most distinctive tactics is accepting failure during development in exchange for speed.

The company favors:

prototype → field → observe → modify → redeploy

rather than:

requirements → specifications → development → testing → certification → production

over many years.

This explains some of Anduril's unusually rapid product development.

For example, Roadrunner was reportedly developed secretly with internal funding for approximately two years before Anduril presented it to the government. 

The strategy has an obvious advantage:

The government sees a functioning product rather than having to fund a speculative technology program.

But it also produces a major risk.

The Wall Street Journal documented a number of Anduril test failures involving autonomous systems, arguing that the company's aggressive development culture creates a tension between speed and reliability. Anduril's position is that frequent failure is an unavoidable component of aggressive testing and iteration. 

That is probably one of the most important counterpoints to the more enthusiastic accounts of Anduril.


7. Build first, create the market second  EXTREME RISK

Another unusual tactic is that Anduril sometimes develops products before there is a large formal government program.

This creates what could be called bottom-up procurement.

Instead of:

Pentagon → requirement → contractor → product

the sequence becomes:

Anduril → prototype → military demonstration → small contract → operational deployment → larger procurement program.

This is particularly visible in products such as Roadrunner and some of Anduril's counter-UAS systems. 

This approach attempts to exploit a weakness of conventional procurement:

government organizations are often unable to specify technologies that do not yet exist.

Luckey has explicitly argued that defense entrepreneurs should sometimes build solutions the Pentagon doesn't know it needs. 


8. Use small contracts as beachheads

Anduril's strategy does not require every initial contract to be enormous.

The company can use a relatively small operational deployment to establish:

  • customer relationships; 
  • operational credibility; 
  • data; 
  • battlefield experience; 
  • integration into existing networks; 
  • evidence for subsequent procurement. 

Once the technology is proven, the contract can expand substantially.

This appears to be one reason Anduril has pursued numerous different military applications rather than betting everything on one large program.

Defense News reported that Anduril's 2024 growth came from a combination of:

  • Collaborative Combat Aircraft; 
  • counter-UAS; 
  • software; 
  • Taiwanese loitering munitions; 
  • autonomous submarines. 

The diversification is strategically important because it reduces dependence on any one weapon program.


9. M&A is another deliberate growth tactic

Anduril has also used acquisitions differently from a traditional conglomerate.

It has acquired companies whose technology, people or military relationships can be integrated into the Anduril architecture.

For example, it acquired portions of Numerica, adding radar, signal processing and command-and-control capabilities to Lattice. 

Anduril's CEO Brian Schimpf reportedly described looking for small companies with:

  • good technology; 
  • military relationships; 
  • growth potential; 

and then using Anduril's resources to accelerate them. 

This creates another version of the platform strategy:

Acquire capability → integrate into Lattice → manufacture through Arsenal → sell as part of the broader Anduril ecosystem.


10. Venture capital is part of the strategy—not merely financing

This is a particularly important difference from conventional defense companies.

Anduril has raised enormous amounts of private capital, including a $1.5 billion Series F used partly to build its manufacturing infrastructure. 

The capital allows Anduril to:

  • fund R&D before government contracts; 
  • absorb failed experiments; 
  • acquire companies; 
  • build factories; 
  • hire engineers; 
  • carry inventory; 
  • pursue large programs before revenue arrives. 

The Los Angeles Times characterizes this as using venture capital to disrupt the defense business model. 

That is potentially the key economic innovation:

venture capital absorbs development risk that traditionally would have been transferred to the government.

In exchange, investors expect enormous growth if Anduril can become a major defense prime.


11. The ultimate objective appears to be becoming a new defense prime

There is an interesting tension in Anduril's rhetoric.

It portrays itself as an alternative to traditional defense primes.

But Luckey has also explicitly said that the objective is to become the next major defense prime. 

So the strategic objective is not necessarily to eliminate the defense-prime business model.

It is to rebuild it around a different technology and economic architecture.

A simplified comparison:

Traditional prime

Anduril model

Government specifies

Company proposes

Government-funded development

Venture-funded development

Bespoke systems

Modular systems

Many subcontractors

More vertical integration

Cost-plus frequently used

Fixed-price emphasis

Long development cycles

Rapid iteration

Small production runs

Mass production objective

Hardware-centric

Software-centric

Platform-specific software

Common Lattice layer

Specialized factories

Flexible manufacturing

Program-by-program

Ecosystem/platform


12. The most interesting recent development: scale changes the economics

The newer publications suggest Anduril is moving from an engineering startup to an industrial company.

That is a major transition.

The Information reports that Anduril expects revenue to reach approximately $4.3 billion in 2026, while simultaneously expecting very large losses because it is investing heavily in factories, employees and new products. 

The thesis is essentially:

Lose money now to build manufacturing capacity that makes the company highly profitable later.

The factory spreads fixed costs over increasingly large production volumes.

That makes Arsenal-1 much more than a factory expansion. It is a bet that defense manufacturing can achieve something closer to automotive/electronics economics.


13. International expansion is part of the model

Anduril is also deliberately building an international customer base.

Its activities include:

  • Australia; 
  • United Kingdom; 
  • Japan; 
  • Taiwan; 
  • South Korea; 
  • NATO; 
  • potentially Poland and other European countries. 

The strategy has several advantages:

  1. Diversifies government customers.
  2. Expands production volume.
  3. Creates additional battlefield feedback.
  4. Builds allied interoperability.
  5. Creates export markets.
  6. Reduces dependence on one Pentagon program.

CB Insights identifies international partnerships as one of the company's major strategic pillars. 

More recently, NATO selected Lattice for its Enhanced Air Command and Control data-platform initiative, providing an example of how the software platform can expand independently of an individual weapon system. 


14. The biggest strategic risks identified by the literature

The publications are considerably more skeptical when discussing economics and execution.

Risk 1 — Customer concentration

Fortune reported in 2026 that CEO Brian Schimpf described Anduril's business as "outrageously concentrated." 

Despite international expansion, the U.S. government remains enormously important.


Risk 2 — Enormous capital requirements

Anduril is no longer simply writing software.

It is building:

  • missile production; 
  • rocket motors; 
  • aircraft factories; 
  • radar manufacturing; 
  • underwater-vehicle facilities; 
  • test ranges; 
  • Arsenal-1. 

That consumes enormous amounts of capital before the resulting production generates revenue.

The Information reports that Anduril remained deeply unprofitable while investing aggressively in factories and expansion. 


Risk 3 — Manufacturing is much harder than software

This may be the most important unanswered question.

It is relatively easy for a software company to go from:

1,000 users → 10,000 users → 100,000 users.

It is much harder to go from:

100 drones → 10,000 drones.

You need:

  • suppliers; 
  • quality control; 
  • tooling; 
  • inventory; 
  • workers; 
  • testing; 
  • logistics; 
  • certification; 
  • repair; 
  • spare parts. 

Anduril is betting Arsenal can solve this.

Reuters reported in 2026 that Arsenal-1 began production of FURY and was expected eventually to employ more than 4,000 people. 


Risk 4 — Rapid iteration can produce unreliable systems

The Wall Street Journal reporting on failures is important because it provides the counterweight to Anduril's own narrative.

The company has experienced:

  • autonomous-system failures; 
  • testing accidents; 
  • aircraft development problems; 
  • operational shortcomings. 

The strategic question is therefore not whether Anduril can develop quickly.

It is whether it can develop quickly without sacrificing military reliability. 


Risk 5 — Lattice creates both opportunity and dependence

Lattice could become a powerful platform.

But the same characteristic that creates competitive advantage can create criticism:

If Lattice becomes the common software layer connecting many weapons, sensors and autonomous systems, Anduril becomes increasingly embedded in the military's infrastructure.

The 2026 academic literature explicitly discusses this in terms of platform lock-in, path dependence and procurement influence. 

That is a more sophisticated criticism than simply saying "Anduril makes weapons."


 


Publications I would read first

If your objective is to understand Anduril as a business rather than simply its weapons, I'd prioritize these:

  1. CB Insights — "Building the autonomous battlefield" — probably the best concise strategic overview of acquisitions, partnerships, Lattice and Arsenal.
  2. Los Angeles Times — Palmer Luckey/Anduril profile — particularly useful for understanding the fundamental business-model philosophy. 
  3. The Information — "Inside Anduril's Big Gamble: An Ohio Weapons Factory" — particularly useful for understanding the economics and risks of Arsenal.
  4. Defense News — Arsenal manufacturing strategy — excellent detail on how the manufacturing model is supposed to work.
  5. Reuters — Anduril's defense-tech expansion — useful independent reporting on actual programs, international expansion and manufacturing.
  6. Wall Street Journal — Anduril failures — important counterweight to the company's promotional narrative.
  7. 2026 Science as Culture paper on "mil-tech oligarchy" — the most useful academic treatment I found of Lattice as a platform and its implications for procurement and military institutions.
  8. Janes — Anduril and the cost-plus model — useful for understanding why Anduril believes conventional defense economics are structurally defective.
  9. Fortune, May 2026 — useful for the current economics, customer concentration, Lattice's emerging role, international expansion and the company's transition toward profitability.

One particularly useful way to study Anduril is to treat it as a three-layer business: Lattice (software/platform) → autonomous products (hardware) → Arsenal (manufacturing platform). The publications make considerably more sense when viewed through that framework.

 

660.5

Warren Buffett Lessons

 

Berkshire Hathaway Lessons

 

Charlie Munger Lessons

Software Lessons from Buffett, Munger, Berkshire Hathaway

 

TOP CONCEPTS SUMMARY FOR B-TO-B 10X SOFTWARE BUSINESS 

 

Link to Berkshire Hathaway, Warren Buffett, Charlie Munger Lessons (Summary by Tom Ingram, password required)  

 

 

TRAPS, BIGGEST MISTAKE:  Miscalculating the underlying economics of a business 

 

- How possibly know what/how to evaluate? ONLY way is LEARNING CURVE, HAVING DONE IT BEFORE. 

 

- FUNDAMENTAL TRUTH OF FOCUSING ON NARROW HIGH VALUE urgent compelling need  

 

*Military history-best at AFTER ACTION REPORTS 

 

* NEVER DO ANYTHING FOR FIRST TIME if possible 

 

- STEWARDSHIP OF O+WNER’S INTERESTS / CAPITAL FIRST (Luke 16:2, parable of unrighteous steward, from 2002 shareholder letter pages 16-21) 

 

- RATIONAL MANAGEMENT AND EXPECATIONS:  15% Compounded annual return is goal.  Has managed 17%+ over 50 years. 

 

- NO SPECULATION! 

 

- VALUE IS PRESENT VALUE OF FUTURE PROFITS – NOTHING ELSE!!! 

 

- If not paying income tax – not interested – not a real, viable company 

 

- DURABLE  BARRIERS TO COMPETITION (SWR GENERALLY BAD) 

 

- WILLINGNESS TO ADMIT MISTAKES, TAKE INPUT, CHANGE 

 

* JP Hogan:  #1 reason military excellence not adopted by commercial.  AAR report example. 

 

* Easy to underestimate.  Business people won’t open kimono, admit weakness 

 

- AVOIDS “TOO HARD” PILE:  TECH, Speculative, turnarounds, industries in trouble 

 

- AVOIDS "fast moving technology which… CANNOT RELIABLY EVALUATE LONG TERM ECONOMICS” 

 

- AVOIDS businesses that only work with SUIPERSTAR CEOs 

 

- TRAP: Single biggest mistake is bad estimate of underlying economics  

 

- Rigorous CEO Selection (normally long term in place CEOs) 

 

- Fair CEO Compensation BASED ON RETURN ON INVESTED CAPITAL 

 

- BOARD effective, interests aligned, non-interference, NO POTTED PLANTS 

 

- HOLDING COMPANY CEO: 

 

   * NON-INTERFERENCE, EXTREME DELEGATION, Authority / Responsibility Matched 

 

   * Willing, COMPETENT to override management when self interest conflict is skewing judgement 

 

- INCENTIVES, PENALTIES EFFECTIVE for board, management, employees  

 

- Only what you understand.   VIGILANT TO NOT GET DRAWN INTO TECH DON’T UNDERSTAND 

 

- DISCIPLINES TO CONTAIN NEGATIVE CONDUCT 

 

- Costs low and going down 

 

- Happy customers – HIGH VALUE CREATES BARRIERS TO COMPETITION 

 

- Happy employees 

 

- Loyal suppliers 

 

- Products continually improving 

 

- GROWTH IMPERATIVE NONSENSE 

 

- NO ACCOUNTING, FINANCIAL WALL STREET, DEAL, BORROWING WASTE OF TIME NONSENSE!!! 

 

- LONG TERM, BUY AND HOLD for 10 years minimum.  NO EXIT PLAN TO SELL TO GREATER FOOL 

 

- Low debt 

 

- ACQUISITIONS:  Rigor, discipline, clear value to owners 

 

- MARGIN OF SAFETY   

 

- Fix or sell under performing businesses immediately (when clear can’t be fixed) 

 

- Not making decisions in front of salespeople / advocates 

 

- Much time spent reading, being up to date, independent, unpressured thinking and judgements 

 

- PATENCE BEYOND BELIEF:  Acquired only 30 whole businesses in 50 years 

 

- INSURANCE INDUSTRY PARALLELS 

 

  * Buffett EMBRACED THE INTANGIBLE / HARDER TO UNDERSTAND when he moved away from textiles into insurance  

 

  * Reinsurance, Business Property, Casualty are B to B 

 

  * INTANGIBLE LIKE SOFTWARE 

 

  * DISCIPLINE REQUIRED (Underwriting willingness to walk is essential.  Reserves - kept conservatively) 

 

  * REQUIRES FOCUS  Wide variety of insurance types – easily distracted 

 

  * Treated as commodity, UNDER ATTENDED by SENIOR MANAGEMENT until TROUBLE 

 

  * VERY LOW BARRIERS TO COMPETITION 

 

  * SUPERSTAR MANAGEMENT is primary barrier to competition (Working for Buffett but RISKY - hard to duplicate when leaders are indispensable) 

 

  * Capacity easily added – JUST ADD MONEY but competitors have same advantage 

 

  * Healthy pricing MARGIN OF SAFETY REQUIRED because nasty surprises are common 

 

- See DEAL SHENANIGANS master list on TIA website 

 

- Possible to TRIAGE QUICKLY TO SUPER HIGH VALUE APPS WITH MOATS AND PREDICTABLE EARNINGS – like Jack Henry, other winners 

 

- FRANCHISE CONCEPT:  “No close substitute in customer’s eyes”  (No urgency, customer DIY, many competitors are swr historical problems) 

 

++++ EXTREME HIGH VALUE for customer possible, GREAT PRICING POWER 

 

+++ No regulation 

 

----- COMPETITORS HAVE SAME ADVANTAGES 

 

+++++ SMALL INCREASES IN COST TO CREATE WHOLE PRODUCT SOLUTION CAN CREATE LOLLAPALOOZA EFFECT ON TOTAL PRICE AND SALES VOLUME 

 

+++ Competitors DISTRACTED BY COMPETING ON TECH, PRODUCT LEADERSHIP 

 

+/- Returns in inflationary times??? 

 

+++++ WHOLE PRODUCT SOLUTION on NARROW, FOCUSED SET OF CUSTOMERS creates EXTREME VALUE and MOATS (High barriers to competition) with STRONG FUNDAMENTAL VALUE 

 

+++ LOW CAPITAL REQUIRED – GREAT RETURNS POSSIBLE 

 

+++++ POSSIBLE TO DEMAND PROFITABILITY FROM YEAR 1 

 

+++ Prices For Good Businesses With Good Long Term Prospects Are Sky-High As Of 2018 Letter.  By mandating profitability from year 1 and these other disciplines, MAY BE POSSIBLE TO PREDICT WINNERS BETTER THAN IN PAST – before price is bid sky high 

 

- Swr lends itself to Munger recommendation to take a simple, basic idea and take it very seriously  

 

   * e.g. Solve competent authority / responsibility / competence problem in swr??? (includes removing perverse incentives, replacing with good) 

 

   * e.g. Solve the narrow focus on super high value solution problem (includes payback triage, scope change discipline) 

 

+++/--  DISCIPLINE NECESSARY:  Swr has not yet evolved economic systems to prevent vice, reward virtue 

 

--- REMEMBER – BUFFETT LOST $13BB ON IBM!!! 

 

+++ Swr industry is generally focus-averse.  TOO EASY TO SAY "WE CAN DO THAT" 

 

++/-- Discipline is unpopular at the moment 

 

+++/--- FIX PROCESS FIRST, Process Reengineering, all required for big-bang implementations in TOO HARD PILE at the moment.  APPEARS SEPARATING PROCESS IMPROVEMENT FROM TECH SOLVES THE PROBLEM 

 

+++/--- AUTHORITY / RESPONSIBILITY / COMPETENCE MATCHING IS ALMOST UNHEARD OF 

 

--- Family, multi-generational businesses not likely 

 

---/+++ EXECUTIVE EXCESSIVE SELF REGARD rampant.  "I know what I need to know", "I don't need to know the details", “Don’t take up my time…” RAMPANT,  must be contained 

 

---/+++ FEATURE COST / BENEFIT TRIAGE almost unheard of 

 

+++/--- Scaling up, economies of scale different from insurance – must be managed closely.  RE-USING LEARNING CURVE, NOT HAVING TO DO EVERYTHING THE FIRST TIME ARE CRITICAL.  Integration to poor customer systems will be huge issue. 

 

++/--  CUSTOMERS FED UP with  immaturity, poor results, HATE SPENDING THE MONEY, ATTEMPTING COST CONTAINMENT but STILL NOT GETTING RESULTS desired / needed, making same mistakes over and over 

 

-----  INDUSTRY MAY BE SO CONTAMINATED – BETTER TO STAY AWAY 

 

   * Compare to investing money in mutual fund before committing (what acquisition focused managers fail to do) 

 

   * TROUBLED INDUSTRY, Insurance Lesson:  NEAR INFINITE SUPPLY, LIMITED DEMAND created extreme price competition to no profits for nearly all competitors 

 

- REMEMBER BASIC ECONOMICS CAN PREDICT MUCH:  Supply up – PRICE DOWN / Demand up - PRICE UP.  Successfully predicted impact of aluminum tariffs, egg price up, gas supply in Europe during Ukraine war.  GETS IN TROUBLE WHEN PREDICTING TOO MUCH, CLAIMING ACCURACY BEYOND BASICS.     

 

- TOUGH CONVERSATIONS SOMETIMES NECESSARY, NEED TO BE DONE SOONER RATHER THAN LATER:  Watch out for the great executive who is succumbing to Alzheimer’s or dementia or some other debilitation.  Happened a few times and Buffett / Munger did not act soon enough.  Directors – speak up! 

 

Sources and References 

 

FROM https://www.berkshirehathaway.com/letters/letters.html. 

CM (CHARLIE MUNGER) CITATIONS BELOW ARE FROM POOR CHARLIE’S ALAMANC, EXPANDED 3RD EDITION 

WB (WARREN BUFFETT) CITATIONS BELOW ARE FROM ANNUAL LETTERS 

BH (BERKSHIRE HATHAWAY) 

 

660

Danaher Lessons

 

Acquisitions Based on Continuous Improvement

 

Amazon Lessons

 

Toyota Lessons

 

RELX Lessons

 

Hands On, Lean, Continuous Improvement Lessons for Software

 

Danaher Model Shows Possible Path to Changing The Software Industry!

 

Obsession With Creating Value For Customer! 

 

Obsession With Better, Faster, Cheaper – Every year – Every Department!

 

Sources: 

  • The Surprising Success of Hands-On Leaders, Harvard Business Review, November 2025 by Scott Cook, Cofounder of Intuit and Nitin Nohria, Former Dean, Harvard Business School
  • HBR Interview with Scott Cook
  • Summary of Lean applied to Software while working for Mitel, being led by a former Danaher executive https://tiainc.net/SkinnY%20Support%20for%20Lean%20Summary.pdf 
    • NOTE:  This failure was likely due to absence of controlling interest and Danaher executive support for Lean / Continuous Improvement
  • Claude:
    • Continuous Improvement contrasted with Lean
    • 30 Year Shareholder Returns for Danaher, Berkshire Hathaway and Fortune 500
  • HBR’s Danaher Case, Updated 2015  [link Harvard Business School]  (This is a purchased case.  Tom’s markup available by Zoom call only).

 

Too many CEOs come up through the ranks of finance, strategy, sales, consulting – NOT OPERATIONS!  They have no confidence in their ability to make operational improvements!

 

This explains much of what I see – executives REGULARLY under attending the hard, necessary PROCESS improvement work needed for major change efforts

 

Danaher, Amazon, Toyota and RELX are the exceptions. Execution, operations effectiveness, walking the talk are fashionable and demanded.

 

Some of the best CEOs create systems and cultures that help everyone excel at getting work done. 

 

WRONG: conventional leadership theory wants CEO’s to avoid being drugged into day-to-day operations. It urges them to focus on vision, strategy, and delegation.

 

These companies demonstrate dramatic improvements, and performance with deeper engagement, obsessing over the customer, designing systems of execution, using experiments to make decisions rather than presuming executives have unfounded superior knowledge. They teach operational two kits in a bed excellence into daily practices.

 

These CEOs are much more engaged in the WHAT and the HOW. The result is a sustained competitive advantage.

 

What CEO does not claim to care about execution? What is the reality that you have seen?

 

RELX is the best performing stock in the four year history of the FTSE 100 index in London. The CEO obsesses over how to measure customer value. He insists on doing it using the customer’s economics! (Tom: I learned, the hard way, that customers don’t pay any attention unless they are using their numbers). IS THE CUSTOMER BETTER OFF WITH OUR PRODUCT? HOW DO WE PROVE IT?

 

These leaders reject metrics that show how the customer benefits the company.

 

Amazon: Bezos built the company by demanding they actually deliver the lowest prices on 1000 items. Fast reliable delivery.  In the early days, Bezos built desks, and pack boxes himself. He determined to keep a hand in the work.

 

PowerPoint presentations are forbidden. Software teams have a great deal of autonomy. Every proposal must be written in six pages or less. (Amazon found written memos to be better at requiring detailed thinking than PowerPoint.). Amazon encourages no title, no holds barred debate. Decisions are made quickly, consider provisional and revised as new information comes in. 

 

Decision rights are close to the front lines.  This is difficult work because it requires leaders to give up control, Redistribute authority, remove approvals and reduce frictions.

 

Toyota:  Decisions are made by testing-not rank or hunch whenever possible. The data makes the decision. Authority comes from the evidence.  This prevents command and control dysfunction.

 

KAIZEN: Know customer and what they want. Observe to understand problems. Root cause analysis by data. Brainstorming. Rapid prototyping.  (the list of things Toyota does well goes on and on and on.)

 

DANAHER: Larry Culp, now leading GE, “we forced division presidents to develop a command of the hows so that they can teach the how. They say come do it with me“. Hands on, in the trenches work is not a phase to outgrow. “You can’t allow yourself to get distracted from the real daily work of the organization” 

 

They strive to be better, faster, cheaper – every year, forever.

 

Failure and invention are inseparable. Failures will happen. You live with it for the upside.

 

A rare skill: the ability to move between altitude and detail!

 

Mission clarity – not micromanagement!  (Making decisions for people)

 

=========

 

ADDITIONAL POINTS FROM SCOTT COOK, FORMER CEO OF INTUIT INTERVIEW WITH HBR:  https://hbr.org/podcast/2025/12/why-great-leaders-focus-on-the-details  

 

MICROMANAGEMENT IS MAKING DECISIONS FOR PEOPLE

 

STRATEGY ALONE WITHOUT ABILITY TO EXECUTE = FAILURE

  • EXECUTION = SYSTEMS AND PROCESSES!!!
  • IN THE DETAILS OF EXECUTION
    • HE GOT TOO DEEP - INTO MICROMANAGEMENT

 

TOYOTA STARTED AS A LOOM MAKER IN RURAL JAPAN IN 1930S

  • FROM 1950 TO 1990 BECAME BEST CAR MAKER IN WORLD
  • TOOK FAILED GM PLANT, MADE IT HIGHEST QUALITY, LOWEST COST IN GM
    • SPECTACULAR CONTROLLED TEST >>> RARE IN BUSINESS

 

STAY CLOSE TO THE WORK - NOT MAKING DECISIONS BUT CONSTANTLY REINFORCING SYSTEMS PROCESSES VALUES PEOPLE

  • WHAT HAPPENS WHEN YOU ARE NOT IN THE ROOM?
    • IF RIGHT THINGS > THIS IS WHAT YOU WANT

 

FOUNDER VS PROF MGMT:  SMALL SAMPLE OF 4 - NO DIFFERENCE

 

OBSESSING OVER CUSTOMER VALUE - MANIC ABOUT DEFINING UP FRONT

  • DO NOT DELEGATE SELECTION, DEFINITION, MEASUREMENT OF CUSTOMER VALUE
    • NOT YOUR PROFITS!!!
  • RELX:  VARIETY OF BUSINESSES - DIFFERS FOR EACH
    • WHAT CUSTOMER VALUES!!!
  • HAVE TO STAY HANDS ON, EVOLVES OVER TIME

 

DICTATING WHEN NECESSARY 

  • BEZOS - NO POWERPOINT
  • AMAZON SOFTWARE 
    • WENT TO "SERVICES MODE"
    • DICTATED CLEAN APIs BETWEEN SERVICES

 

BEFORE YOU DICTATE >>> EXPERIMENT TO MAKE SURE DICTATING THE RIGHT THINGS!!!

  • MANY COMPANIES NOT WELL EQUIPED
    • NEED 
      • SETS OF TEST CUSTOMERS
      • CONTROL GROUP - CAN KEEP TESTS VALID
      • TOP COMPANIES REALLY GOOD
        • BEZOS - INVESTED IN TESTING SYSTEMS
          • MAKE EASY
          • MAKE HIGH VOLUME
  • MOST OF THE TIME, DECISIONS CAN BE TESTED!!!
    • IT IS CULTURE AND WILLINGNESS
    • RARE EXCEPTIONS WHEN CAN'T
    • EVEN DECISIONS BY SENIOR EXECUTIVES

 

TOYOTA STORY > TESTING TEAM LEAD SOLUTION AND PLANT MGR SOLUTION AT SAME TIME

  • EVERY OTHER BUSINESS > PLANT MGR DECISION PREVAILS

 

TEACH THE TOOLKIT OF THEIR ORGANIZATION

  • PARTICIPATE WITH TEAMS - DON'T MAKE DECISIONS FOR THEM - COACH THEM ON RIGHT TOOLKIT

 

CONTINUOUS IMPROVEMENT

  • RELX - CEO INSPIRING
    • BETTER FASTER CHEAPER EVERY YEAR
      • EVERY DEPT!!!
  • DOES NOT HAPPEN NATURALLY >>> TOO MANY EXCUSES

 

AT COMPANY ALL MGMT MEETINGS

  • PRESENTATIONS ARE "HOWs" FROM TEAM LEADERS
    • 10 MINUTE SESSIONS PICKED BY SENIOR EXECS
  • NOT SPEECHES FROM CEO!!!

 

WHAT NEEDS YOUR ATTENTION NOW?  Hard question – always have big time demands

  • STRATEGY - SHOULD BE RARE

 

CEO TIME SPLIT

  • WHAT GETS SHORT SHRIFT IS MOSTLY "HOWs" OF GETTING GREAT WORK DONE
  • SPEND MORE TIME WITH CUSTOMERS – they care about
    • ENGINEERS who build their products
    • CUSTOMER SERVICE who fixes problems
    • NOT CEO OPINIONS!
  • SPEND LESS TIME AT CONFERENCES, WITH PRESS

 

ACID TESTS TO KNOW GETTING IT RIGHT

  • RIGHT THINGS HAPPEN WHEN CEO / EXECS NOT IN THE ROOM
  • MANY EXPERIEMENTS, SOME FAILURES
  • VALIDATED RESULTS FROM CUSTOMERS
  • VALIDATED BARRIERS TO COMPETITION, MARGINS SUPPORTING ROIC YARDSTICK
  • HOW CEO / EXECUTIVE TIME SPENT

 

HBR DANAHER CASE OBSERVATIONS BY TOM

 

BEFORE:  NOTE HOW THEY EVOLVED FROM VINYL AND RUBBER MFG TO VERY HIGH DEBT / LEVERAGE / TAX DRIVEN ACQUISITIONS, BRUTAL (NECESSARY) COST CUTS AND DIVESTITURES, HOSTILE AND FRIENDLY,  STAYING CLOSE TO INDUSTRIAL PRODUCTS, EXITED AUTO INDUSTRY

 

NOW:  SCIENTIFIC, TECHNICAL INSTRUMENTS, DENTAL, LIFE SCIENCES, DIAGNOSTICS, ENVIRONMENTAL, TEST, MEASUREMENT, INDUSTRIAL TECHNOLOGIES

  • LESS CYCLICAL
  • BETTER MARGINS
  • BETTER BARRIERS TO COMPETITION
  • LESS DEBT, DEBT RISK

 

SEE EXHIBIT 6 - OPERATING INCOME FOR HIGHER VALUE SEGMENTS OUTSTANDING

 

STARTS WITH CUSTOMER (MARKET / INDUSTRY ATTRACTIVENESS - NOT AQUISITIONS FOR SALE!)

 

MORE AGRESSIVE, LESS SELECTIVE THAN BUFFETT IN ACQUISITIONS

 

DBS OFFICE - CONTRAST TO PMO INEFFECTIVENESS!!!

 

DIFFERS FROM BUFFETT WHO DOES NOT HAVE A "STABLE OF EXECS WHO HAVE BEEN PROMISED CHANCE TO RUN A COMPANY"

 

STILL PUSHING ON GROWTH, EARNINGS GROWTH.  LESS ABLE TO RESIST WALL STREET THAN BUFFETT

 

ONLY IN ACQUISITIONS, BUSINESSES WHERE DBS CAN BE CONSISTENTLY APPLIED (INDUSTRIAL ONLY, NO CONSUMER, NO TECH??? - ASK DOUG DICKERSON ABOUT INET... WAS IT DIVESTED?

 

GOT THROUGH 2008/9 IN GOOD SHAPE - ABLE TO TAKE ADVANTAGE

 

DEB TO EQUITY:  EXHIBT 13, 2002 TO 2010, JUST 20-30% OF EQUITY.  CHERRY PICKED?

 

GROWTH 1/3 ORGANIC, 2/3 ACQUISITION 2002 TO 2010

 

CUSTOMER VALUE FOCUS:  NOT STRESSED BY HBR.  SEE EXHIBIT 4 - LISTENING TO CUSTOMER PROMINENT.  ALSO RADIOMETER STORY PAGE 12 GETS TO CUSTOMER FOCUS

 

 

 

SUMMARY OF TOTAL RETURNS, LAST 30 YEARS, 9/22/2026 from CLAUDE

  • S&P 500
  • BERKSHIRE HATHAWAY
  • DANAHER 

(Close approximations)

Company/Index

30-Year Annualized Return (CAGR)

30-Year Cumulative Total Return

$10,000 Invested 30 Years Ago Would Be Worth

Danaher (DHR)

~14.9%

~6,397%

~$649,700

Berkshire Hathaway (BRK.A)

~11.2%

~2,250%

~$235,000

S&P 500

~10.3–10.9%

~1,770–1,900%

~$187,000–$200,000

 

 

 

============

 

Summary of Lean applied to Software while working for Mitel, being led by a former Danaher executive https://tiainc.net/SkinnY%20Support%20for%20Lean%20Summary.pdf 

 

============

KEY DIFFERENCES BETWEEN LEAN AND CONTINUOUS IMPROVEMENT:

 

Continuous improvement is a general philosophy: always look for small, ongoing ways to make processes, products, or services better. It isn't tied to any one method. Kaizen (the Japanese term), PDCA cycles, Six Sigma, and Total Quality Management are all forms of it.

 

Lean is a specific management system, rooted in the Toyota Production System, that has continuous improvement built in. Its central goal is to maximize customer value while eliminating waste, meaning anything the customer wouldn't pay for, such as waiting, overproduction, defects, and unnecessary motion or inventory. Lean comes with its own principles and tools, including value stream mapping, 5S, pull systems, just-in-time production, and standard work.

 

The key differences

  • Scope: Continuous improvement is a mindset that can be applied anywhere. Lean is a defined system with a particular focus on waste and flow.
  • Focus: Continuous improvement aims at getting better in general. Lean aims at getting better by removing waste and creating value for the customer.
  • Toolkit: Continuous improvement has no required tools. Lean has a well-known set.
  • Relationship: Lean includes continuous improvement, but continuous improvement can exist without being Lean. A company using Six Sigma to reduce defects is doing continuous improvement without necessarily being Lean.

 

This ties to the Cook and Nohria article, where kaizen (continuous improvement) is one practice, but Danaher's consistency comes from the broader system around it. As the authors put it, you can't just borrow kaizen and expect the same results. GE under Culp adopted Lean specifically, with continuous improvement as its engine.

 

659

 

Article in process:  AI Economic Model Emerging As Cloud Computing GOOD WSJ Aug 2026

658

Barriers To Competition

AI Barriers to Competition, Moats, Where Major Companies Will Focus

 

“Good Enough” AI at commodity prices will become the widespread norm.

 

Useful Article - but Omits Vertical Focus 

 

Code Generation Emerging as Best Niche – Barriers Possible

 

WSJ 07 18 2026 by C. Mims

657

No Code / Low Code / User Programming / 

 

Code Generators

 

Lessons for AI

7/13/2026 Session Summary:  AI Code Generators, Low Code, No Code Solutions

 

The Big Takeaway:  Best work for senior software people is LARGE, COMPLEX, HIGH VALUE APPLICATIONS / WORKFLOWS.  Integration, Architecture, Complexity which AI is not ready for yet. 

 

Avoid the small, simple software apps where a junior person can get the job done with AI’s help. 

 

See Tom’s Success Story for using AI Code generation on a large, complex project.  Includes process improvement, use cases, architecture, WBS (Work Breakdown Structure), Gantt Chart schedule and Agile development Click Here for Full Case Details #3.41

 

Summary:

POSITIVES / GOOD NEWS

  • Users who self serve are CLOSE TO THE WORK. Understand priorities, what works, what doesn’t
    • Tools like Excel have displaced unneeded complexity such as “cubes” in business intelligence, big data by being close to the work
  • Users who self serve are MOTIVATED.  Tom Note:  Best results come from 
    • Users who built the app themselves
    • When power users and executives have significant financial incentives tied to measured outcomes.  When no one will put the project in a comp plan, you know you are low on the priority / value / impact list
  • Much faster than traditional development – but must include the disciplines below
  • Tools with STRONG INTEGRATION are likely to have biggest impact, lower cost
  • Rows / Columns of data, Visual Records, Graphics make apps much more accessible 
  • Gartner study claims 60% of business processes change every 6 months, 18% change monthly, 4% change weekly (have not found study confirming)
  • 50% of IT spending is controlled outside of IT in top performing companies click here (password required)

NEGATIVES / BAD NEWS

  • AI-generated code must be reviewed / understood by humans.  This need goes up exponentially with application complexity
  • Be SKEPTICAL of vendor claims. 
    • Many products do not live up to the hype. Get rolled out with MARGINAL to POOR user engagement.
  • PE and other investors are ANGRY at LACK COST SAVINGS through HEADCOUNT REDUCTIONS. Not materializing as they were promised.

UNDECIDED / TOO EARLY TO TELL

  • AGENTS: huge hype, potential promise, but need to go slow. SOLID PAYBACK, implementations, few and far between.
  • SECURITY remains a big concern, but MYTHOS has earned GOVERNMENT approval for high security situations
  • OPEN SOURCE:  Good article arguing that Open Source just continues all problems of code based solutions.  Context:  Federal Government.  CLICK (password required) 

DOs 

  • PSPATICDAUBB ACRONYM to remember key items
    • Payback
    • Scope / Change Management
    • Process
    • Architecture
    • Testing
    • Integration
    • Cyber security
    • Documentation
    • Audit Trail
    • Upgrade path/costs
    • Back up/disaster recovery
    • Break / Fix
  • Focus on High value, high payback, (10x) high impact applications. The effort will survive inevitable setbacks and overruns if problem / solution is high enough.
  • Enterprise Apps with GOOD LOW / NO CODE TOOLS WORK BEST. 
    • Framework / data of existing app gives users structure, context – gets better outcomes
    • Third party tools usually FAR BETTER than native enterprise app tools – but companies often get acquired and go away
  • GO at a SENSIBLE PACE. Failures to date have been largely from trying too much, too quickly, without taking time to have fundamentals in place.
  • DATA: focus on HUMAN INSPECTED data. Cannot take AI inspected data as true
  • USE AI where PERSONAL PRODUCTIVITY IS BENEFICIAL, especially presentation, summary, analysis, but REMEMBER that COMPLEX WORKFLOW with MANY USERS is ENTIRELY DIFFERENT
  • TESTING needs to include rigorous NEGATIVE OUTCOMES due to AI immaturity at present
  • CONTROL WHAT AI IS ACTUALLY being USED
  • DEALING WITH THE IT DEPARTMENT:
    • IT often gets runaround, ignored.  Will say "It is your problem..." for good reasons…
    • Big risk of becoming last on IT's priority list
    • Working constructively with IT on the PPTICDAUBB issues mentioned above – can win their cooperation.  Also helpful:
      • Contingency plans if key people gone.  
      • Preventing performance under load / high volume problems
      • Prevent bad vendor choices
    • Focus on very high value problems / solutions.  Senior management will REQUIRE IT TO COOPERATE.  (If senior management won’t – you are headed for big trouble)
    • Co-Funding by Business Unit, placing a code gen / no code / low code person in every department or some alternative to the current IT to business unit dysfunction is probably required

DON’Ts

  • Trust vendors to deliver “BEST MODEL” (big topic)
  • Allow vendors to set the pace (big topic) 
  • TOO MANY, LOW VALUE FEATURE REQUESTS:  Tom Note:  I’ve been caught by this trap several times.  We want to please our users, we see value / payback in these requests – but EXECUTIVES JUST WON’T FUND THEM. 
    • In some cases, executives think no code / low code means “can do in-house,  don’t need to pay a consultant”
  • Get distracted, pulled away from the business plan, priorities, and quarterly objectives
  • The FRANKENAPP Problem:  A motivated super user creates a complex app ONLY HE OR SHE UNDERSTANDS.  App becomes mission critical, layoffs happen – no one around who can fix or upgrade.  Worse it becomes a JOB SECURITY crutch 
  • Allow NEW TOOLS, unless there is a clear value
  • Allow UNSOUND PRACTICES
    • E. G. Assuming anything in AGENT or AI INHERITS to be TRUE
  • ALLOW TIME PRESSURE to CIRCUMVENT DISCIPLINE. (Big topic)
    • Possible exception: no/low code solutions properly controlled may meet time demand safely
  • OUTPACE HUMAN WORK / TRUST

See Previous Newsletter Items For Successes, Lessons, Detail:  497 (Lawson, BEST), 493 (Correlation), 488 (Reseller Success, Very Good) 457 (Seibel, Mixed) 450 (Progress, Huge Win for Shareholders) and 654 (Summary of code generator / no code / low code history and lessons)

 

 

656

 

Article in process

 

6/2/2026 Wall Street Journal:  AI Big Players Now Giving Away AI Capacity or Selling Cheap

 

OpenAI Anthropic Profit Projections Ridiculous WSJ 04 06 2026

 

ChatGPT Now Selling Ads on Free Versions – Citation pending

 

655

Software Industry Financial Lunacy Being Laid Bare

 

Software Will Get Cheaper, More Complex with Less Support

 

 

6/2/2026 Wall Street Journal Finds Software Company Revenues Artificially High Due to Covid, DEBT NOW BEING DOWNGRADED.

 

AI Also Threatening Software Company Viability

 

TOM PREDICTION:  Our research found in 2022 that the software industry was, as a whole, unprofitable and high debt / bad balance sheets were everywhere.

 

Under this stress, software companies will “Add Features and Cut the Price”.  (Michael Porter concept).

 

EXPECT SOFTWARE TO GET CHEAPER and MORE COMPLEX – with LESS SUPPORT and many software companies becoming irrelevant or just going away.

 

Summary from article: 

 

  • COVID-19 pumped up software company sales due to remote work. 

 

  • Software firms borrowed heavily against those revenues, but had to borrow PRIVATE DEBT. Their financial statements were so outlandish that they could not issue public debt.

 

  • NOTE: Wall Street firms have gotten very good at schmoozing customers to buy private debt. WSJ found four major “debt sales companies“ are particularly good, accounting for 25% of these debt sales.

 

  • Revenues have now DROPPED, but the DEBT REMAINS.

 

  • 6% of software company debt has now been marked down over 20% by credit agencies. (healthcare debt is the only sector that is higher). 

 

  • LIKELY to get WORSE. Will take six months plus for the opaqueness to work its way through to public debt/Bond markets.

 

  • An ETF holding SaaS stocks was down 24% while the rest of the stock market has boomed.  It has come back up by 13%. 

 

  • SOFTWARE accounts for 17% of institutional investors debt. 

 

See Wall Street Journal, 6/6/2026

 

654

No Code / Low Code / User Programming / 

 

Code Generators

 

Lessons for AI

Summary of No Code, Low Code Solutions, Code Generators Last 45 Years

 

Big Lessons for AI:  WE HAVE SEEN CODE GENERATION GOOD / BAD BEFORE!

 

 

See Newsletter Items For Successes, Lessons:  497 (Lawson, BEST), 493 (Correlation), 488 (Reseller Success, Very Good) 457 (Seibol, Mixed) 450 (Progress, Huge Win for Shareholders)

 

Following is a summary of points and key software with a link to the full detailed article at bottom – password required.

  • Defense Contractor effective use of CODE GENERATORS as part of WBS.  Included AUDIT TRAILS, Testing, Integration, Cyber Security required by government Click Here for Full Case Details #3.41
  • Notable No Code / Low Code Software Packages 
    • Mainframe Era:  MAPPER, IBM Tools, SAS, SPSS.  BREAKTHROUGH:  Columns and rows of data.  MAPPER History: Click Here.  Critical concepts Tom first saw in 1981:  User programming, Rows and Columns (Spreadsheet) for defining, understanding problems, Visible Record Data (instead of complex databases - execs and users understand), Quick Prototyping, Hostility from conventional IT.
    • Minicomputer Era:  DEC Datatrieve
    • PC Era:  Lotus 123
    • Client / Server Era:  DBase by Ashton Tate, Dataflex, Lotus Notes, MS Access...
    • Scripting Language, Program by Demonstration
    • Web Era:  BREAKTHROUGH - Netscape/Firefox, easily available to all, big leap in user skills, business knowledge much closer to the problem...
      • BREAKTHROUGH:  SharePoint, Rise of files and images as data elements with Web Sites as container for jobs (Old systems can only handle numbers and letters well)
      • BREAKTRHOUGH:  Videos Available to Demonstrate on Web
    • Salesforce: 12+ user programming add-ons to core CRM applications
    • Cognos, Nintex,  Oracle's BI product, Tableau, Qlik
  • DEVASTATING IMPACT Of The "Franken-App" Problem For  A  Used Car Parts Client:  20+ acquisitions, all allowed to develop no-code solutions as desired with no documentation.  Tribal knowledge gone due to layoffs, NO MARGIN to pay for effective offshoring / rework of critical apps
    • MS Access and other user built solutions have a long history of being developed by a hot shot, evolving through multiple people, no documentation, extremely complex, becomes mission critical to the department and the hot shot leaves.  
    • IT wants nothing to do with this.  "It is your problem..." for good reason.  
    • Solution:  Best to recreate from scratch with good tools, documentation, backup, etc.  User departments MUST budget for these expenses.  
    • Big risk of becoming last on IT's priority list. 
  • Databases, SQL got much better easier to use.  Report writers and business intelligence tools are an important part.
  • REMEMBER, third-party no-code tools tend to be way better than enterprise software vendors tools, BUT they tend to get acquire and stop all R&D for other platforms.
  • "Why Can't We Run the Business on Excel?"  (Visible Record Data - execs and users understand), Quick Prototyping, Hostility from conventional IT.
  • Gartner study shows 60% of business processes change every 6 months, 18% change monthly, 4% change weekly (Mike Fitzmorris, Nintex, 7/21/15, DFWSPUG Talk, have not found study confirming)
  • See "Bowling Alley of No-Code Solutions" for Five TIA Clients
  • 50% of IT Spending is Controlled Outside of IT in Top Performing Companies CLICK HERE (password required)
  • Best Fit, Fastest Results seen in Process Improvement and Business Intelligence
  • How To Work With the IT Department:
    • IT is getting bypassed all the time.  Hard to compete with "Just give me a credit card and you can start tomorrow"
    • Co-Funding by Business Unit:
    • IT provides 101 capabilities at no charge.  Identify 202, 303, 404 projects and payback.  Business Unit contributes e.g. 10% of the payback
  • Client Side Development (Javascript, JQuery) Emerging as Far Better than Modifying Server Code
  • Big Data, Analysis of Large Data Sets, Business Intelligence:  Cubes Complexity, Cost No Longer Needed, Justified.  Simpler Tools Like Excel Better Because Close to the Work.
  • "Citizen Developer" Concept (Gartner):  Success Stories at Displacing SAP, Oracle, Big ERP at Flowserve, Fisher Controls, Others  Click Here (password required)
  • Case for No Code Solutions instead of Open Source:  Good article arguing that Open Source just continues all problems of code based solutions.  Context:  Federal Government.  CLICK (password required)
  • History of User Programming / No-Code Solutions 2010 (still useful points video)
  • Tips for Coping with IT Department Reaction, Mistakes to Avoid:
    • Fundamental mistakes that result in "I told you so", IT taking control later:  Poor documentation, no backup / disaster recovery / contingency plans for maintaining, improving solution if key people gone, compatibility problems, performance under load problems, cannot grow to handle increased volume, bad vendor choices.
    • Failing the Cost / Benefit / High Impact, Rapid Results Think-through, Trying to Do Too Much with Too Little:  No-code solutions should shoot for a 10X clear payback over five years and settle for no less than 5X.  The full life cycle cost needs to be planned for.

Link to full article, requires password

 

653

 

Who Gets the Margin?

 

Economics of AI Don’t Make Sense

 

Article in Process.

 

“…subscribers of the top [AI] models…  pay 1/40th of the actual cost.”  From Wall Street Journal, 6/17/2026 by Holman Jenkins

653

 

Who Gets the Margin?

 

Summary of “Who Gets The Margin?” Newsletter Articles

 

New Addition:  Sam Insull, the 1920s utility baron who electrified the economy but ended up broke.  From Wall Street Journal, 6/17/2026 by Holman Jenkins

652

 

Executives Not Competent to Lead Major Operations Improvements

 

Article in Process.

  • Per Robert Schaffer Rapid Results by Schaffer & Ashkenas, Copyright 2005:  “Business cultures often undervalue operations – cast it as low in status, boring, unglamorous. Many senior managers schooled in finance, strategy, deal making and marketing simply don’t know much about operations – leaving it to the engineers A SURPRISINGLY LARGE NUMBER OF PEOPLE REACH SENIOR MANAGEMENT POSITIONS WITHOUT EXTENSIVE EXPERIENCE IN GETTING THEIR TROOPS TO TACKLE VERY DIFFICULT CHALLENGES SUCCESSFULLY. (from page 56)

651

 

AI Slop Code Consequences

 

What It Takes for Production Quality Code

 

Good AI / Bad AI

 

Only 13% Of AI Code / AI Output Has a Clear Owner Responsible When Things Go Wrong

 

87% Lacks Clear Owner for Security, Integration, Testing, Production Support, Compliance / Audits, Data Integrity and Business Continuity 

 

This is What Separates Solid, Complex Workflows and Applications from “AI Slop Code”

 

(Thanks to Ron Giblin for contributing this summary)

 

See the CloudBees 2026 survey on AI-generated code https://www.cloudbees.com/lp/2026-state-of-code-abundance-report

 

Good Uses for Quick AI Code”

  • Prototypes (very helpful in defining user requirements)
  • Testing
  • Documentation
  • Repetitive Coding
  • Developer Productivity (when inputs and outputs can be clearly defined)

Trouble When Undisciplined Code is Relied On:

  • More Bugs
  • More Testing Burden
  • More Technical Debt
  • More Cloud Usage / Token Costs / Other Costs
  • More Unclear Ownership

 

650

 

Rapidly Entering New Markets

 

Shortening Defense Procurement Process

 

Understanding AI Costs

 

Startups and Bigger Firms Using AI More Than Mid-Size Firms

 

(Thanks to Ron Giblin for contributing these two summaries)

 

Dr. M. Ray Perryman recently cited federal business trend data showing that AI use is much higher in larger firms.  Companies with at least 250 employees are far more likely to be using AI than smaller firms.  See article for checklist / guidelines https://www.census.gov/library/stories/2026/05/ai-use-businesses.html

 

Startups Using AI To Rapidly Enter New Markets

 

See HubSpot’s research: https://www.hubspot.com/startups/ai/ai-stats-for-startups  Includes detail on thinking through startup issues.

 

TOM NOTE:  Watch the Defense Industry startups – especially in Silicon Valley.  These startups have found a way to DRAMATICALLY SHORTEN THE GOVERNMENT PROCUREMENT PROCESS.  

 

Major defense contractors sell programming / software / product development programs to the government – a long, arduous, problem prone process.  Startups use AI to rapidly create complete products and offer them to the U.S. government.  The government says “yes or no”, DRAMATICALLY SHORTENING PROCUREMENT TIMES! 

 

Thanks to Jim Sherrill for this insight.  

 

Understanding AI / Token Costs

 

See Microsoft article tips on where to watch for costs.  https://learn.microsoft.com/en-us/microsoft-copilot-studio/analytics-cost-savings 

  • Cost per task
  • Cost per customer interaction
  • Cost per resolved issue
  • Cost per generated feature
  • Cost per workflow
  • Cost per error
  • Cost per business outcome

 

649

 

Sloppy AI Code

 

Bugs

 

Outages

 

Security Breaches

 

Technical Debt

 

Precedent:  Off-Shore Programming

 

5/23/2026  Key AI Architects of Open Source Agent, Coding Tools Sound Alarms

Coding Crisis Coming – Inevitable

FORECAST:  Companies will realize bad software is being created without the expected cost savings.

Claude Code:  BIG Criticism for bugs, screen flickering, memory hogging, feature creep

 

Wall Street Journal article describes likely problems identified by Mario Zechner and Armin Ronacher – architects behind OpenClaw open source coding tool.

 

Inevitable Consequences

  • Buggy Software
  • Service Outages
  • Security Vulnerable – WILL BE ATTACKED
  • Mountains of Technical Debt
  • Github, other useful tools clogged with bad code

 

Effective Uses:

  • Testing
  • Quick prototypes
  • Human oversight essential – MODELS EASILY GET OFF TRACK

 

Weaknesses:

  • BIG, COMPLEX, INTEGRATED SYSTEMS
    • Laying off experienced people PERMANENTLY REMOVES KNOWLEDGE that AI is not trained for, cannot replicate

Marginal Uses

  • Agents in larger enterprises

 

See https://www.wsj.com/tech/ai/vibe-coding-slop-ai-tools-e6a99394?mod=Searchresults&pos=2&page=1 

 

648

 

Low Barriers to Competition

 

Deal Fever, Perverse Incentives

 

Disastrous Price Assumption

 

FrankenApp Risk

 

Global Crossing Precedent for AI Valuation Lunacy 

$55 billion valuation down 90% in four years during tech bust, Bankrupt in five years from JUST ONE BAD PRICE ASSUMPTION IN A SPREADSHEET.  

Forecast:  Token Price Down 90-99% in Five Years, 470 Million Coders instead of Current 47 Million Coders

China’s DeepSeek Price Cut by 75%.  OpenAI and Anthropic Cost 15X as Much

Tom:  

  1. “Narrow Focus on Solving a Big Customer Problem With High Margins and Barriers to Competition will be Everything.”
  2. “This will cause more ‘FrankenApp’ disfunction (users will code convoluted apps only they can use, protecting their jobs.)”

Extracts from 6/1/2026 WSJ, Andy Kessler

At the time, undersea T-1 lines (which transmit data at 1.5 million bits per second) cost between $10,000 and $20,000 a month. Their AC-1 undersea fiber could handle 40 gigabits per second, or more than 10,000 T-1 lines, which could easily generate more than $100 million a month. Spreadsheets flashed green!  

 

 Salomon Brothers and Merrill Lynch took Global Crossing public in 1998, raising $400 million. The stock peaked at $55 billion in value. Global Crossing borrowed heavily to build out 100,000 miles of fiber.  

 

 At the time, undersea T-1 lines (which transmit data at 1.5 million bits per second) cost between $10,000 and $20,000 a month. Their AC-1 undersea fiber could handle 40 gigabits per second, or more than 10,000 T-1 lines, which could easily generate more than $100 million a month. Spreadsheets flashed green!   

 

 Salomon Brothers and Merrill Lynch took Global Crossing public in 1998, raising $400 million. The stock peaked at $55 billion in value. Global Crossing borrowed heavily to build out 100,000 miles of fiber.  

 But it worked only with $10,000-a-month pricing. The world isn’t static. AT&T and MCI created a joint venture to lay their own undersea cables. Prices plummeted. By 2002 the cost of undersea T-1 lines dropped to $1,000 a month, down more than 90%. Spreadsheets flashed red.  

After reporting fake revenue that Arthur Andersen blessed, Global Crossing filed for bankruptcy.

See https://www.wsj.com/opinion/the-hallucinatory-ai-math-075f1112?mod=author_content_page_1_pos_1 

 

647

 

Fix The Process First!

 

Who Gets the Margin?

 

Verify Demand Actually Exists

 

Durable Barriers to Competition

 

Focused on Right Questions

 

Ignore Lesser Distractions

 

Nov / Dec 2025 Three Solid Essentials for Effective AI Use, Havard Business Review

 

TITLE:  The Genl AI Playbook for Organizations by Andy Wu et. al.

 

SEE CHART ON PAGE 129. TOM note:  the three items on far right of the chart match my thinking exactly. I’ll paraphrase.

  1. FIX THE PROCESS FIRST!!!  BIG FAILURES come from simply throwing new tech and dollars at large scale, complex process inefficiencies.
  2. MAKE SURE CUSTOMERS NEED/WANT what you are producing and WILL PAY a margin that will produce above average returns.
  3. MAKE SURE YOU GET THE MARGINS created by the efficiencies of AI. This requires durable barriers to competition and doing things fundamentally differently than your competitors. You must also guard against your supply chain, gaining bargaining power.

Map tasks, focus on COST OF ERRORS, and TYPE OF KNOWLEDGE REQUIRED.  

 

Embrace AI where it is already useful. 

 

MISDIRECTED, WRONG QUESTIONS: is AI perfect enough? Are we moving fast enough with AI transformation? What are competitors doing with AI?

 

BETTER QUESTIONS: Where can AI create a durable competitive advantage? How can I use AI differently/better than competitors? What are AI implications, upstream and downstream in my value chain?

 

WHO GETS THE MARGIN?  Do you reap the efficiency GAINS or are they negated by competitors? Does your SUPPLY CHAIN gain leverage for lower prices or better terms?

 

PRECEDENT: the PC Transition from command line MS DOS to windows GUI. Took 10 years to mature.  Did not significantly alter large process / enterprise applications.

 

DISTRACTION of AI PERFECTION: The real focus needs to be on effectiveness compared to current work methods.

 

CAUTIONS, WATCH OUT FOR:

  • AI focused ENTRANTS with dramatically, lower cost basis, particularly in labor.
  • CUSTOMERS, and SUPPLIERS, using AI against you, undercutting your bargaining power.

SOME OPPORTUNITIES

  • When your competitors are paralyzed by hallucinations, imperfect intelligence, etc.
  • When you have proprietary data
  • When you have unique people, processes, culture/values. TOM note: Geoffrey Moore describes this as the barriers to competition created by WHOLE PRODUCT SOLUTIONS.

AI will not magically improve your P&L. It will require intentional use and durable competitive advantage.

 

See article for other contributors, including NYU Stern, school of business.  HBR Reprint R2506K

 

646

 

AI Cost / Benefit Viability

 

Who Gets The Margin?

 

Vertical Application Focus Most Likely Winner 

 

Commodity Cautions

 

Long Term Strategy Focus Instead of Reactive Short Term

 

 

9/19/2023 Unlikely BIG AI Investments Will Make a Profit, Havard Business Review:  

 

HBR’s Andy Wu Explains Extreme Challenges 

 

Basic economics of generative AI are being overlooked.

 

  • SAS model does not apply. Variable cost of delivering generative AI to an AI user is NOT zero. Paid subscriptions and higher usage pricing will be essential. Companies are doing this already.
  • CORE. TECHNOLOGY is open source and most data is PUBLICLY AVAILABLE, which will result in LOW BARRIERS to ENTRY, and many competitors
    • (TOM note: WHO GETS THE MARGIN? The customer? A winner in the supply chain while all others get commodity margins? Key question championed by Geoff MOORE and Clayton Christensen.)
  • VERTICAL focus notes:  Predicts vertical AI models and applications will be widely known, extremely competitive
  • WINNERS are still not clear
  • MAJORS ARE NOT in business of SELLING MODEL itself. Upside coming from integration to their applications and selling cloud services.  (Tom NOTE: AI becomes one more feature for cost/benefit analysis and effective or ineffective implementation) 
  • MONEY WILL NOT BE MADE on AI itself. (TOM note: will be made where high application value is created with barriers competition)
  • INVEST/PLAN for AI over next 10 years. Not next year.
  • COMPANIES HAVE INCREASINGLY COMPLEX PORTFOLIOS of DATA and enterprise applications. Integration has been poor. AI will not be a silver bullet fix.
  • Early WINNER EXAMPLE??? Bloomberg partnered, created Bloomberg GPT for financial tasks.  Critical nature of financial insight made developing their own LLM worthwhile.  (TOM note: WU does not identify specific use cases and payback. Whenever this is absent, I strongly question the ROI.) 

 

See https://www.library.hbs.edu/working-knowledge/how-will-the-tech-titans-behind-chat-gpt-bard-and-llama-make-money?utm_source=chatgpt.com 

645

 

Enterprise AI Success Story

 

Agent Success

 

Code Generator Success

 

Good Processes to Start From

 

Needed Time to Capture Right Data for AI Prediction

 

Baby Steps, Quick Hit Benefits before Advanced AI Usage 

 

(See key lessons from 6/8/2026 Zoom session at bottom) 

 

6/8/2026 An Enterprise AI Success Case by LogistixIQ and Chris Cameron, SVP*

 

BEFORE AI: Four dispatchers could handle just under 500 Trucks Of Fracking Sand Per Shift

 

AFTER AI: Four Dispatchers Expected to Handle 1000 Trucks Per Shift

 

Competitor Benchmark:  Chris tells me he knows of a competitor using 8 dispatchers to handle less than 500 trucks per shift. 

 

Chris’s team is looking at DOUBLED OUTPUT for same labor cost – possibly 4X OUTPUT compared to a competitor!

 

 

Chris Cameron, SVP of logistics IQ, shared some amazing results managing about 30,000 truck shipments per month, seven days a week, 24 hours a day. 

 

A TIMING AGENT allows the trail of several hundred trucks to act as storage! Significantly reducing storage, handling, and other costs.

 

The TIMING AGENT also estimates turnaround time to optimize sand supply chain for each well.

 

A RISK AGENT continuously ranks wells at risk of going down for lack of sand, GREATLY REDUCING chances of well going down. AI is able to handle the complex calculations far quicker than humans. Provides continuous well risk management rather than sporadic. 

 

TECHNICAL FOUNDATION is a custom built core system in REACT with a MySQL database. Chris uses Windsurf for AI coding.

 

Chris led the charge to put the AI on top of the course system in less than five months.

 

Chris will be discussing this case with our Senior Software Working Group at 6 p.m. Central time on Monday, June 8, 2026.  Both in person and by Zoom.

 

Email me if you would like to join us.  No cost.

 

======== Key Takeaways, Success Tips from 6/8/2026 Working Group session by Chris Cameron=========

 

  1. Had good basic in-house developed system to start from.
    1. Similar to job shop manufacturing software.  Created a “load of sand”, built out its route then executed route with good measures
  2. Process was understood, strongly embedded in systems
    1. Human Office processes were immature – all over the place.  One of the first things Chris fixed.
    2. Chris had to GET THE PROCESSES SUFFICIENTLY RIGHT to provide GOOD DATA FOR AI to learn from.  AI CAN’T DO MUCH WITHOUT THIS DATA.
  3. REDUCING TIME WASTED on idle trucks was biggest payback.
  4. TOOK BABY STEPS TO GET TO AI USE
    1. Initial quick hit improvement used basic formula to flag P.O.s nearly out of money – huge benefit for trivial cost.  NO ONE CARED THAT IT WAS NOT USING AI!
    2. Used AI to begin predicting truck routes that would waste time!
    3. Used AI to improve Invoice / P.O. / Loading Ticket reconciliation.  Increased “untouched orders” from 50% to 91%!  Only 9% now require human review.
    4. Next Advanced AI Use:  Telling drivers how fast to drive to minimize wait times.
    5. AI now alerts human if a truck is not moving.
  5. BIG GAIN from TIMING AGENT:  Each day has 8 or 9 jobs going simultaneously (overwhelming to dispatchers).    Agent ranks the top two or three needing immediate human attention. 
  6. Eventually AI will fully dispatch the trucks (currently at 1,000/day).
  7. CHRIS’S TIPS
    1. Start small, move ahead, don’t wait for perfect processes / data.
    2. Produced a Minimum Viable Product in a few weeks and began using it.
    3. Organized work into two week sprints – team got used to producing tangible benefits every two weeks – without a scrum master!
    4. PAY FOR THE ESSENTIALS:  e.g. Chris paid for a full week of Decision Tree analysis – well worth it.

 

 

 

644

 

Effective Allocation of Capital

 

Product Development: Work Hard to Find Low Capital Niches

 

Careful – MUST Still Have Barriers to Competition

 

 

5/12/2026 Wall Street Journal Shows 99.99% probability big investments in AI will produce poor stockholder returns 

 

The magnificent seven got where they are by producing huge returns on low invested capital.  

 

They are abandoning that model, investing huge amounts on AI. 

 

100 years of detailed US stock market returns show with near certainty that asset-light companies produce better returns than asset-heavy companies.  

 

History shows that asset-heavy company stock prices do great during booms but lose all the gains when the tide goes out.

 

The very few winners will win big, society will benefit and customers will get advanced capabilities at a bargain price, but the vast majority of this invested capital will produce poor returns.

 

Tom: Our research shows repeatedly that the right question to ask is “WHO GETS THE MARGIN?“  (Who benefits from the productivity gain?) 

 

When a productivity gain is easily duplicated by numerous competitors, the margin ALWAYS goes to the END CUSTOMER and a few select niches in the value chain.

 

Our recommendation is to focus on those niches where barriers to competition can be created and a strong margin can be sustained.

 

See Wall Street Journal Article https://www.wsj.com/finance/investing/leading-stocks-are-losing-their-low-asset-edge-4a2885c3?mod=author_content_page_1_pos_3  

 

643

Effective Allocation of Capital

 

Polar Opposites Provide Contrast

 

Warren Buffett vs. Softbank

 

5/10/2026:  Economist Article Contrasts Two Extremes in Ownership / Investing

 

Implications for Software Product Development, Software Jobs

 

Buffet / Berkshire Hathaway Approach: “Low risk, low debt, big cash reserves, strong barriers to competition, predictable profits on moderate to low invested capital, happy customers, happy employees, consistent solid returns for shareholders, holding good companies for a long time.  Buys companies based on present value of future profits.”

 

Softbank* Approach:  “Invest in any tech that is gut level promising, especially AI, regardless of price.  Extreme debt.  Cash from operations nowhere near able to pay debt.  One in 10 big wins is all that is needed.  Extreme AI investment means Softbank could easily disintegrate.  (Softbank’s demise has been predicted many times.)

 

Tom Comments on Impact for Software Products, Jobs:  Be aware that there is a school thought for software / tech products which is very close to the Buffett approach:  Insisting on profitability from the start, only pursue niches where we solve a big problem for customers, barriers to competition exist resulting in strong margins on low invested capital.  

 

I am repelled by the hype, unmet promises, volatility, lack of concern for customers and employees in the Softbank approach.  This is a complex topic – contact us for details.

 

* Softbank founded by Masayoshi Son, initially an Asian Telecom, has grown into an enormous tech speculative investor based in Japan.  Softbank did well by bringing proven tech trends from U.S. to Asia.

642

AI Lessons from Life Insurance Software

Lessons from North American Life, Midland National Life, Globe Life, AIG Life (now Coforge), U.S. Life

 

Article in process – 

 

Will be subject of AI Software Product Development Working Group Session

641

 

Big Win for AI with Process Improvement

 

Enterprise Software Product Saves Brain and Heart Tissue, Reduces Cost and Doctor Time Waste

 

 

 

Success By Viz.Ai Using AI To Create More Effective Pathways (Processes) 

 https://www.wsj.com/opinion/the-algorithm-will-see-you-now-e8cc8b93?mod=Searchresults&pos=1&page=1  Wall Street Journal article by Andy Kessler 4/20/2026.

 

VALUE CREATED

  • Reduced time to treatment in a stroke by over an hour. Time is brain.
  • AUTOMATED APPROXIMATELY 90% of DOCTOR WORK before seeing patient ( both administrative and low-level decision-making). Frees up doctor for higher level work, dramatically, helping with current doctor shortage.
  • IMPROVED QUALITY of CARE, HUMAN LIFE example:  Can now synthesize a 900 page electronic health record to essential, “action needed “essence, significantly increasing the chance of finding disease before symptoms occur.

 

BEFORE

  • Messy workflow, 13 different decisions and handoffs for stroke
  • 56 decisions / handoffs for thickening of heart muscles.
  • PROCESS TRIGGERS:  Initially triggered by the tech (imaging results processed by AI)

 

AFTER

  • Shortened initial treatment by up to 88 minutes
  • Reduced variance in treatment times per patient by two hours and seven minutes.
  • Faster and more consistent treatment for all patients. 
  • Moved to electronic healthcare record as BETTER PROCESS TRIGGER. Now triggers 90% of workflows. 
  • Now in 2000 US hospitals covering 230, million people and 35 European hospitals. Growing 40% per year.

 

SUCCESS LESSONS

  • INITIAL RESISTANCE:  Initially marketed/perceived as just helping the radiologist, neurosurgeon and doctors read scans.
  • Resistance reduced dramatically when all realized had to fix the entire workflow FULLY UNDERSTAND PROBLEM
  • 20% is detecting disease. 80% is fixing workflow through large bureaucracy.
  • WHOLE PRODUCT SOLUTION To SALES CHALLENGES: Appears selling single disease solution to hospitals was a problem. Expanded to 55 diseases/pathways.  35 pathways came from partners.
  • AI Agents are part of solution. 
  • WORKFLOW DESIGN: started by turning hospital rules into SOFTWARE.

 

COMPETITORS doing similar work include AI.DOC and OpenEvidence. 

 

640

Blackberry Revied by Narrow Niche Focus

 

Right Out of Playbook for Crossing the Chasm, Other Books Advocating Niche Focus

5/2/2026 WSJ article about Blackberry Returning to Profitability Through Niche Focus

Transitioned from handheld to Auto Computer Real Time Operating System niche (RTOS) then to additional niches in medical devices, industrial automation and robotics

NOTE:  TIA’s research in 2024 classified Blackberry as a (huge) ROIC loser (Return on Invested Capital):  

$9 Billion in Capital Tied Up for 9 Years produced a $7 Billion Lost

The news is it has now produced an annual profit for first time in 20+ years. 

 

A big SUCCESS STORY in turnaround through NICHE FOCUS (so far…)

 

Their RTOS product is now found in 275 million cars.  It provides the plumbing that allows key driver, assistance:  Collision warnings, blindspot, notifications, adaptive, cruise control, pedestrian detection, steering correction…  It is designed to never, ever fail and moving on to additional niches in autos, factory floor, hospitals, surgical robots, etc.

 

Some hard work, some luck resulted in finding this great niche. The gigantic blackberry failure drew all attention, leaving this business unit alone to work on solving a big problem for a KEY customer. The desperate circumstances of Blackberry also meant they had no choice but to find a better business model.

 

BREAKTHROUGH came over a beer with a key Customer, Audi’s engineering chief. He said “we are moving our infotainment business to Google but here is what we need for the next generation of cars...“.  The Blackberry unit Head listened and a new, profitable niche focus was born.

 

See article at https://www.wsj.com/tech/blackberry-qnx-software-cars-bf2a2280?mod=tech_lead_story 

 

639

Vertical Software Firms More Stable, Less Vulnerable

4/26/2026  AI-Driven Severe Downturn in Software Company Stocks Shows Industry (Vertical) Focus Holds Up Best Under Stress / Pressure WSJ - see

https://www.wsj.com/finance/investing/wall-street-is-sorting-software-companies-into-winners-and-losers-e44fe73b?mod=Searchresults&pos=2&page=1 

 

Loans to vertical software companies have only been discounted by 4.2%

 

Software Engineering firms discounted 16.3%

 

Horizontal firms by 8.8% 

 

Cybersecurity by 5.3%

 

Inherent advantages of vertical / industry-focused software firms:

 

  1. Customer focused instead of tech focused
  2. Solves, bigger, higher value problems for customers
  3. Higher perceived value by customers due to WHOLE PRODUCT SOLUTIONS
  4. Higher switching costs
  5. Much higher barriers to competition
  6. AI cannot easily replace
  7. More likely to focus on profit, mid to long-term value for owners, happy employees, happy customers
  8. Less likely to be damaged by lunatic speculation, growth at all costs, acquisitions, excessive CEO pay

 

638

 

History of No-Code / Low-Code Solutions

 

Relevance to AI

 

Code Generators, No-Code / Low-Code Solutions, Citizen Developers, User Programming …

 

History, successes, failures, over-optimistic predictions shed light on use of AI to create code.

 

Click here to see article in process  (password required)

 

 

637

Skunkworks

 

Sybase

 

Salesforce

 

VMware

 

Symbol 

 

Short Success Stories

Skunkworks, 1974 by, Kelly Johnson and Ben Rich. 14 rules and RIOT ACT resulted, spectacular, unequaled, defense results over 50 years. Most projects completed in 14 months to two years:

  • First US jet fighter, 200 mph faster than prop driven, working in five months
  • First mach 2
  • The stealth fighter which created a 25 year advantage over Soviets and contributed to their downfall
  • Stealth technology behind the B2 bomber
  • The U2
  • The blackbird, mach 3, SR 71
  • Best air to air missile initial test in history
  • $200 million plus refunded to government is unneeded, unused or impractical


Sybase, 2007 to 2010. Escape Velocity by Geoffrey Moore, page 126. narrow focus on two niches took company from market cap of $2.2 billion to acquisition of for 5.8 billion DURING 2008/2009 GREAT RECESSION.

Salesforce 10 X value creation by reducing enterprise software installation and operating costs versus legacy apps. Escape Velocity, Geoffrey Moore, page 150.

Salesforce results record mixed. BREAKTHROUGH (I think) parentheses: making it dramatically easier for management to get control of sales leads and sales people. Arguably, a net negative for sales people.

VMware 10 X reduction in the cost of data center provisioning and consolidation of unused computing capacity. Page 152, Escape Velocity by Geoffrey Moore.

SYMBOL technologies, 10 X breakthrough, DEVELOP HORIZONTALLY, SELL VERTICALLY with ruggedness of computers for mobile needs.2003 to 2006, page 155, Escape Velocity by Geoffrey Moore

636

 

Bessembinder Findings

 

Only 3.4% of Public Stocks Created Net Wealth in Last 100 Years

 

Corroborates TIA Research

 

95% of public firms don’t outperform cash.  50% lose money per finance professor Hendrik Bessembinder of Arizona State University. Wall Street Journal, 4/9/2022 by Jason Zweig

 

“Most of the return of the stock market over time comes from a few high-performing ‘superstocks.’  More than 95% of all stocks, over their lifetime as public companies, collectively, don’t even outperform cash, and more than half deliver negative returns”

 

Corroborates 2024 TIA Research finding:  “89% of public software companies fail to outperform cash in a mutual fund over 10 years.”

 

REMEMBER:  Only 1-2% of Tech Startups Become Public Companies or are Acquired  “Only the best become public companies” may be an illusion.

 

3/23/2026  BESSEMBINDER CITED AGAIN in Wall Street Journal 

“Research by Hendrik Bessembinder  indicates that almost all the wealth creation from equity investing was produced by a tiny fraction of individual stocks.  

 

For the 100 years since 1926, the U.S. stock market has returned about 10% annually. But only 3.4% of the stocks were responsible for all the wealth creation.  

 

The rest of the stocks combined didn’t produce net gains that exceeded the amount that could be earned from investing in 30-day U.S. Treasury bills. And more than half the stocks in the market lost money.  * The generous average return from the market came from a tiny handful of stocks.  An indexing strategy works because it captures these rare winners.”

 

635

Percent of Services Matters

 

Barriers to Competition

 

Whole Product Solution

(AI’s Impact:  Early 2026 is showing BIG drops in major software company share prices.  We are looking a WHY and what to do…)

 

When Services are Less Than 30% of Revenue, Software Companies are Vulnerable, Unstable

 

Features, Tech Leadership Alone Are LOW BARRIERS TO COMPETITION

 

Your competitors will duplicate your features, cut prices and squeeze your margins

 

Services at 30% or greater shows focus on High Value for customer, WHOLE PRODUCT SOLUTIONS and  creates barriers to competition.

 

Dassault, a $6 Billion French software giant has sales declines and a 50% drop in stock price.  (FYI:  I have some negative experience with Dassault products – all techno hype and no regard for effective implementation.)  Services are 10% of revenues.

 

Major company % of revenue from services per ChatGPT on 2/15/2026:

  • Adobe:  2%
  • Salesforce:  6-7%
  • Oracle:  9%
  • SAP:  13%
  • Microsoft: 2.8%

 

Our research from 2021 shows that software companies producing a long term Return on Invested Capital over 10% have between a 70/30 and 30/70 software to services mix. 

 

BOTTOM LINE:  Competing on tech / product features alone leads to very very rare short term wins, but it is a slippery slope.  Much better to focus on creating extreme high value for customers (a whole product solution to an urgent compelling need) with a mix of product, services and partners.

 

Sources:  WSJ 2/12/2026, TIA Research from 2026 software company stock price drops, TIA Research Update #3 2021, 49 Firms Studied 

 

634

Who Gets the Margin?

 

Containing the Failings of Human Nature

 

Owner Interests Ahead of Personal Interest

 

Rational Management

“Exciting technology or healthcare stocks end up earning lower returns than companies making bottle caps or toilet paper.”

 

Robert Haugen, University of California, Irvine, wrote a rigorous book “The New Finance” (out of print.)

 

He argued that “the pathetically inefficient market doesn’t seem to have a clue as to what is going on.” 

 

His analysis of data from 1928 to 1992 concluded “the risk-return trade-off is truly negative,”  

 

Why? Less-volatile stocks are priced too cheaply “because they are boring.” That’s why buyers can earn higher returns in the long run. 

 

Source:  Wall Street Journal, 2/7/2026, THE INTELLIGENT INVESTOR | JASON ZWEIG

 

633

Who Gets the Margin?

Who Will Reap the PROFITS from New, Big Technology???

 

THE GRETZKY METHOD:  Always asking “where will the profits be?” 

 

Good Competition / Bad Competition

 

Clayton Christensen did a great service by insisting we REMEMBER to think through VALUE CHAIN PROFITS.  See example page 155. 

 

He walks us through the PC Revolution and shift to non-proprietary hardware as example.

 

He shows WHO WON THE BIGGEST MARGINS AND ROI FOR OWNERS – by far:

  • Microsoft
  • Intel
  • Applied Materials
  • Disk drive / head designers and manufacturers 

 

The rest devolved to either commodity status or simply went away.  Think of all the big names that are now no longer relevant:  IBM, DEC, Compaq, Dell, Data General, Wang, Computer Land, Business Land…

 

WHO DOESN’T GET THE MARGIN:  Feature based competition is easily copied – results in nothing but reduced margins for all competitors – and customer reaps the value / margin.

 

Bad competition allows others to reap the value of improvements. 

 

Good competition allows you to reap the value.

 

Gretzky taught us to “always ask where is the puck going?”  AI fans pay attention.  We need to be asking “who will get the margin?”

 

Source: Innovator’s Solution by Clayton Christensen, 2003

 

632

Authority / Responsibility / Competence

 

Profitable from Day One

 

Not Chasing Every Dollar of Revenue

 

Scope, Requirements, Change Control

 

Narrow Focus on Extreme High Value For Customer

 

Work Breakdown Structure Discipline Leads to Accurate Status

 

Cost / Margins

 

Sales Execution Excellence

 

Urgent Compelling Need

Big Lessons from the Navy, Long Career in Software.  Interview with Ted Puchacz January 2026.

 

Ted was in the Navy in Intelligence and had a big career in the software industry leading sales teams for IBM, Nashua, UCCEL, Computer Associates, Siebel and Axiom.  Ted and I discussed some key things he has seen improve software outcomes:

 

  1. Matching authority, responsibility and competence is by far the most important item in Ted’s view.

 

  1. Profitable from Day One.  Stay away from the nonsense of giving Techies infinite amounts of money and letting them waste it.

 

  1. On Scope, Requirements, Change Control: IMPLEMENT THE BASE SOFTWARE FIRST!!!  Then tightly scrutinize the requested exceptions.  Yes, it will mean saying “no” to some lower-level and mid-level people.  Yes, they will have to change their business processes to match the software.  This is a big issue because Ted worked in the Heyday of Siebel when it pioneered CRM as a standalone application.  Siebel was a darling of the stock market and boomed in the tech-boom of the late 1990s-early 2000s, but was severely criticized for extreme cost and complexity of implementations and upgrades.  This translated often into serious customers’ dissatisfaction. The single biggest thing Ted saw in getting over that problem was NOT SAYING YES TO CUSTOMIZATIONS IMMEDIATELY. DO NOT CHASE EVERY DOLLAR OF REVENUE. AGAIN, IMPLEMENT THE BASE SOFTWARE FIRST.

 

  1. Executive engagement, buy-in, control and effective status reporting. Ted made the point that usually you only have to do STATUS REPORT DISCIPLINE a few times. You demand accurate status reporting and take immediate action on  shortfall.  People that can’t meet their commitments are removed.  People that hide problems are removed.   People that accurately report status are rewarded.  “Your status reports will get dramatically more accurate – real quickly.  You won’t have to enforce it again. But you must be willing to demand performance.”

 

Cost / Margins, Cutting the Team in Half:  A Computer Associates, Charlie Wong story on managing and optimizing the cost of technical delivery: Charlie Wong was the controversial head of Computer Associates. There is good news and bad news about Charlie Wong’s approach, Ted:  “Charlie taught me that on every big software project if you start out with a team of 20, a few months in you can cut it by half, because half of the people are doing the primary work and you’re beginning to understand who they are.  In another few months you will be able to cut that in half again, down to 5 people or so.  It will be very clear who is performing / contributing and who is just hanging on.  

 

  1. Sales Execution Excellence Story: Ted had a big win for Nashua earlier in his career competing against Memorex selling disk subsystems. Though Memorex’s product was cheaper, Ted took his territory from $60,000 in revenue per year to $1.2 million per year.  He did it by meeting a key compelling need for state governments.  On a 5-year lease, Nahsua’s price was competitive with Memorex’s and winning while competing with IBM, Memorex, 3M & BASF.

 

Ted was able to overcome an Urgent Compelling Need – state governments cannot commit to a lease  for more than one year. Ted structured a rebate of discount rebate where, if the state government renewed the following year, they got the discounted price and the total 5-year cost was attractive.  If they could only pay for one year and did not renew (which never happened) they simply paid for the difference between list price and the discount price.  

 

631

Containing Human Failings

 

Leadership Includes Industry Expertise

 

Quality of People Doing the Work

 

Costs / Margins Earned for High Value Work

 

Scope, Change Control, Requirements

 

Narrow Focus On High Value For The Customer, ROI

 

Urgent Compelling Need

 

Finding the High Value ROI

 

Fix the Process First

 

Prevent Technology Fascination

 

Costs / Margins / Risk

 

Fix the Sales Problem

How Big Consulting Firms Increase Software Project Success Rate from National Average of 35% to As High As 80%**

 

 

My research starting in 1998, the Standish Group 2019 Survey, and numerous other sources put National success rates for business software projects around 35% as promised.  The majority of projects in large companies actually deliver only 42% of promised features and functions.

 

The big firms have a number of advantages.  Mark, a veteran of Financial Accounting and Reporting Niche Practice from both Deloitte and KPMG, shared how - summary below.

 

The big consulting firms have made a business of doing a better job for their customers – and this is why they get paid well.  Mark’s view over a long career is that projects were consistently well above industry averages and often reached 80% or better as promised, defined as the  “Customer would do it again.”  While these measures are imperfect and subjective, the fundamental disciplines and things that the big firms do are clear, straightforward and repeatable by smaller firms. Importantly, these outcomes were most consistent in projects with clear executive sponsorship, enforceable scope control and alignment between business and technology owners.

 

Ethics and Partner structure:  Mark and most of his co-workers came from Audit practices where ethics and integrity were everything.  This structure causes partners to share reputation, financial, and governance risk for delivery quality. This enforces an accountability that we do not see in many other sectors performing software projects.

 

Industry Expertise:  The big firms nearly always include serious executive industry expertise on their teams.  It  costs more, but adds credibility, wins sales and prevents repeating dumb mistakes

 

Good to Great People:  The big consulting firms charge significantly more than many consulting and services companies to do the same work.  A significant portion of that cost is due a consistently higher baseline of trained, coached, reviewed, and mentored talent, with strong internal quality controls on who is staffed to client work..

 

Big Firms work with Bigger Customers, willing to pay for Quality Help.  Again we see that price, margins, profits are interrelated with the ability to staff, govern, and protect quality delivery through disciplined controls.

 

 Scope, Change Control, Requirements:  Mark learned the hard way that the person writing the written agreement for Scope must have experience over 20+ projects.  He also learned the sales process never gets enough detail for adequate scope and change management. THERE MUST BE A POST-SALE EXTENSIVE WRITTEN AGREEMENT ON WHAT WILL ACTUALLY BE DELIVERED.  

 

He also found you must REIGN IN THE ENTHUSIASM.  There is so much excitement when a sale is closed that team members go off in wrong directions.

 

Narrow focus on high value for the customer, High ROI:  Mark’s practices were primarily financial reporting using Oracle and other financial reporting tools.

 

ON FINDING THE HIGH ROI:  Mark found that some 20% of projects the customer was nearly desperate, at risk of failing an audit, failing disclosure rules or the financial reporting problem was so big the customer did not want to take the time to do formal ROI calculations.  They wanted a solution now. This is defined as a UCN (Urgent Compelling Need) in the management literature. When the customer has to act.

 

In the other 80% of clients, the ROI came from the impact of the accuracy of financial reports and how quickly  management got it. Labor savings alone will not justify the project.  The anecdotal impact of mistakes that management had made in the past because of late or poor financial reporting produced all the ROI justification needed to proceed with these projects.  (Tom:  I have used this “cost of past mistakes” approach to ROI very successfully.)

 

Fix the Process First:  The big firms all have standard best practice process templates.  Mark’s practice required in the contract that the best practice process would be implemented prior to implementing the technology.  This might be the single biggest thing the big firms do right.  They insist, even over lower level objections, that the client process be fixed before applying the technology.  Mark noted that front line low level and mid level people often resist changing the process where senior management universally supports getting to best practice processes.

 

Business Result First Prevents Technology Fascination:  The narrow focus of Mark’s niche lent itself to keeping “Technology for Technology’s Sake” pressures at bay. This is another of the key issues that the big firms do right.  

 

Scope, Change, Risk Control #2 - the “Deal Review Board” was a critical control mechanism.  It was rigorous, escalated reviews based on dollar volume and insisted on a profit margin commensurate with risk. Margins in the high 20s to 40% range were typical for approval, with higher-risk or less-defined projects requiring either higher margins, additional controls, or outright rejection.  (Tom:  My research shows that 38-40% margins are necessary to keep a firm continuously improving and able to weather storms.  Below 38% I have seen a downward spiral of poor people, poor systems, unhappy customers and declining profits.

 

The Sales Problem:  The big firms are noted for spending a substantial portion of their effective cost structure on sales activity when partner selling time is included.   They don’t call it that, and the Partners’ business cards don’t say, “Salesperson”, but that is what they do. This is another essential to the big firms’ success.  They spend the money to build a pipeline of work so they can pick the good projects with good margins, rather than taking anything they can sell just to meet payroll.  (Tom: I worked for a former big firm partner that helped me understand why big firms spend 1/3 of revenues on sales / marketing.  I watched IBM do this and conducted my own research showing that at less than 28% spent on sales / marketing, software companies tend to decline.)

** (Tom:  The improvement to 80% success rate was not provided by Mark.  It is a “guesstimate” based on my research and a long software career.  It is not rigorous or statistically proven.  My point is that the big consulting firms do things that dramatically improve client outcomes – we can all learn from these lessons.)

 

 

Authority / Responsibility / Competence Matching

Notes On Authority / Responsibility / Competence Matching, January 2026

 

Key seems to be placing authority as close as possible to the decisions / work.

 

  • SOUTHWEST AIRLINES vs. American Airlines:  When asked what he thought was the most important contrast in his 25 year career at Southwest, Doug E., a veteran of reservations, customer service and cargo cited “the authority – at the front line – to do the right thing for the customer and the company – and knowing your supervisors will back you up.”  Doug cited numerous examples such as a customer with a death in the family would receive a free rebooking.  A caution:  “I eventually moved out of customer service because of “nonsense” demands such as “I want a refund because I was stuck in a center seat.”
  • SUBMARINE CREW PERFORMANCE, DRAMATIC IMPROVEMENT:  In the book “Turn The Ship Around”  U.S. Navy Captain David Marquet tells how his subs achieved big, measurable improvement:
  •           U.S.S. Idaho’s preparedness rating improved from the bottom (3) to almost top of fleet (nearly 9 on 10 point scale)
  •           U.S.S. Santa Fe improved retention rate by 300% and “Officers to Command” by 900%
  •           Accomplished without adding people, money, or better equipment—only better leadership practices

His most important factor?  Vesting the Chiefs (front line supervisors) with authority and requiring them to take ownership responsibility for improving outcomes.  See the book – a great read.

  • Authority to Match Responsibility, Without Interference (“broad latitude to accomplish goals as he/she sees fit) was cited as the #1 most important factor in job satisfaction and retention for middle and later career people.  (Source:  Wall Street Journal, Jan 2026, citation pending.)
  • Israeli Army Authority Lessons:  Person per person, many military historians believe Israel is the best.  As a small nation of 10 million, surrounded by 200 million Arab people, survival has depended on getting the highest possible performance from each person.  Among the top techniques Israel uses is pushing authority downward – to the point of best information – to match with responsibility.  Competence is matched by effective training.  The amazing victories of the Israeli military and the supporting details can be found in Certain Victory, by General Robert H. Scales, 1994, Warrior: An Autobiography by Arial Sharon, 2001 and The Sword of Freedom by Yossi Cohen, 2025.

 

 

630

Urgent Compelling Need

 

Narrow Focus On High Value For The Customer, ROI

 

Cost / Margins to Pay for Good People

 

FGIC Problem (Financial Guys In Charge) 

Lessons from Convex Super Mini Computers:  Convex came to dominate the Super Mini Computer segment in the 1980s and 1990s with more than 500 units installed worldwide.

 

Convex took a significant number of sales away from Cray Computers because it found a niche, “Half the power of a Cray Supercomputer for one-tenth the cost”.

 

Convex is cited by Regis McKenna in Relationship marketing for defining and dominating a segment (super mini computers).  Cray was too established in super computers and DEC was too strong in mini computers.  

 

A friend who worked for Convex for some 15 years mentioned some of the things that made Convex:

 

  1. Focus: Did one thing very well.
  2. Earned High Margins by solving important problems for customers.  These high margins were necessary to afford superior tools and people.
    1. For example, Brad explained that even a simple software product requires dozens of servers, lab area, sophisticated software tools and competent people to simply do compatibility testing, stress testing, cyber testing, etc.  A super computer required much, much more for R&D, testing, user applications, etc.  Convex earned the margins necessary to pay for good tools, systems and people.
  3. Decline in Leadership when acquired by HP: Initially, the leadership of HP were engineers that came up the ranks – but that changed. Carly Fiorina brought in a shift to “the financial guys in charge”. (Tom: note that Charlie Munger and Warren Buffett have been openly critical of Carly Fiorina’s leadership.) Brad saw the decline and eventual demise of the Convex Unit.

 

629

Fix The Process First 

 

Containing Human Failings

 

Requirements, Scope, Change Control

 

Use Cases

 

Exceptions Control

 

Authority / Responsibility / Competence Matched

 

Quality of “Doers”

 

Narrow Focus On High Value For The Customer, ROI

 

Cost / Margins

“What I learned from a $20 million disaster.”

 

A colleague with 35 years experience in the software business (Oracle, PwC, Bearing Point) told me about the hard lessons learned leading a huge failure for CBRE (the largest real estate and property management in the U.S.)  The application was hand-held facility management.

 

  1. Best Practice Process Abused: “Process is oversold and corrupted by self-interest of vendors and consultants.  The vendors and consultants in this effort claimed to have ‘best practice’ processes.  In reality, these processes were BIASED TOWARD SELLING THEIR TECHNOLOGY.  The end was an unworkable system that users rejected  - and complete failure.” (Tom: I have seen this abuse directly and personally.)
  2. Solution: Start from the As-Is Process.  ”Do the work to document the As-Is Process and understand why it exists. Then compare it to the alleged best practice. Where it makes sense, adopt best practice.  Where it doesn’t, figure out how to adapt to what is working for the organization. 
  3. THE KEY IS TO FIX THE PROCESS FIRST, BEFORE BUYING THE TECHNOLOGY!
  4. Focus on the 80% Use Cases that accomplish the primary work. Have written agreements up front  that prevent the distraction of trying to handle all exceptions.
  5. Containing Human Failings: “One of my biggest takeaways was that politics and self-advancement played a key role in this devastating failure.  Numerous people were continually pestering upper management with ‘I have a better solution, better software’, in attempts to advance their own interests, rather than what was best for the company.”
  6. Tom Comment: Authority, Responsibility and Competence Matching are some of the keys that prevent these problems. Berkshire Hathaway (Warren Buffet and Charlie Munger) have pioneered incentive systems and other controls that drastically reduce the problems caused by the fallen nature of human beings.  My colleague and I reviewed my 1998 research showing that this fundamental problem was one of the big four issues that has to be corrected to achieve effective software outcomes.
  7. Quality of Young People Decreasing Due to AI Dependance: My colleague hires younger people and interns on a regular basis and has noted a distressing trend.  Younger and entry-level jobs are fewer and the people who apply for them are too dependent on AI and the internet for their solutions.  They lack understanding of the fundamentals.  Solutions to this problem are a much bigger issue, but a good beginning is to focus on a narrow set of customers where you produce extreme high value and can afford to hire the best of available people and train them is necessary.

 

 

628

 

Leadership

 

Authority to Say “No”

 

“Mechanical / Electrical Engineer In Charge” Problem

 

Requirements 

 

Scope / Change Control

 

Architecture

 

Authority / Responsibility / Competence Matched

 

Sales Problems, Pressure to Sell

 

Selling to Board Members, CEOs

 

Technical Execution Excellence

 

Commander’s Intent

 

Ross F, former Texas Instruments Program Manager and Software Team Lead for 16 Years, Led 12+ Programs / Projects (Classified and Unclassified).  Huge Win Selling His Own Tech Company for Millions, Other Big Software Rolls / Wins, Retired Lieutenant Colonel, U.S. Army Tank Commander in Dessert Storm

 

Ross and I met in 2025 and 2026 to discuss his big lessons from and long career in Defense and related software projects.  My summary of key takeaways:

 

  1. Software leadership must have the authority to say NO.  The leadership team must also include Mechanical and Electrical Engineer and Sales heads, but WATCH OUT FOR THE “EE / ME / Sales In Charge” problem.  Software is intangible where EE and ME disciplines are largely based on physical elements (case, circuits, semiconductors, etc.)  Sales also does not possess the skills / discipline to execute software.  These leaders, in practice, will not enforce the necessary disciplines such as Requirements / Scope / Change control.   Unfortunately, the norm is that software leadership gets overridden and the software is late, over budget and not as promised – time after time after time.  (Tom: I have also seen these problems to devastating effect.  Time after time after time.)

 

  1. Architecture:  Product must be designed with TWICE the memory and processing power for initial requirements or the product will become obsolete.  

 

  1. Competence: The Military REQUIRES people to come up through ranks - know the jobs they are supervising.  Military is also good at matching authority to responsibility.  (Tom and Ross agree - the Commercial side often allows “politicized incompetence”, the equivalent of the “POLITCAL GENERALS PROBLEM” that was so devastating in the U.S. Civil War.)

 

 

  1. "Just Get The Product Out The Door" Problem:  We see this everywhere.  Must find the balance.  (Tom:  Yes, we must sell and make a profit – but there are solutions.  Keep reading.)

 

  1. End Customer (Government) Problems That Must Be Managed:  "Do not even attempt a pilot unless you have an order in hand and 25% down.  The customer can change their mind and withdraw.  Requiring a written agreement and a deposit forces at least some basis to manage scope and enforce change control.   It forces sales, manufacturing and engineering to all focus on a contained scope - instead of all the possible bells and whistles”

 

  1. Sales Problems That Must Be Managed:  “You cannot allow sales people to define the product.  Software and engineering must engage and have authority.  I found the software lead had to travel with sales to see the customer and be present for all commitments.  I learned this the hard way but once I insisted on traveling, it became easy.

 

  1. Burning All Cash to Deliver Pilot:  “This might be devastating, or it might be OK if you know you will get follow on work.”

 

  1. Get the first win and do a great job!  “Board members and CEOs network with each other.  Do good work and others will hear about it.”

 

  1. “Clear Commander’s Intent is Everything.  The best possible value for the customer is job one.  Only pursue business with adequate margins in our focus area.  Scope, change control, development, testing, clear disciplines that are enforced.  Architecture designed for enhancements and upgrades, honest sales representations, honest status reports, PSEUDOCODE…  These are all examples of communicating what the military calls Commander’s Intent.”

 

Tom:  I require “Fix The Process First”.  How can you have a clear commander’s intent when you buy technology and throw it at a broken process?

 

  1. “Why I Require Developers to Write Pseudocode First:  It forces them to write (in English!) the concepts and flow of the solution.  It leads to good documentation, testing, security and ability to fix and enhance software use of Artificial Intelligence.”

 

 

629

Authority / Responsibility / Competence Matched

 

Saying “NO” to Bad Business

 

Not Chasing Every Dollar of Revenue

 

Extreme High Value for Client First

 

Margins To Do It Right

 

Fanatic Controls on Risk, Margins, Merit of People

 

Quality of Leadership, People

 

Strategy

 

Leadership Accountable

 

Competence, Hard Truth and Results – Not Good Intentions

 

Disciplined Process for Selling, Contracting and Execution

Lessons From Big Consulting – The Best of the Best:  Accenture

 

2024/2025  I’ve had the chance to learn from some long term Accenture people.  No firm gets it perfect every time (we discuss a failure below) but Accenture’s record of growth, profitability and happy customers deserves study.  Some lessons:

 

MANAGING DIRECTORS ARE IN CHARGE! 

  • They have both sales and delivery authority / responsibility / competence matched. 
  • They are incented and penalized when they sell good or bad business. 
  • My contact SAW HALF DOZEN EXECUTIVES DEMOTED, LOSE BONUSES OR LEAVE THE FIRM because they sold bad projects.  
  • They regularly say NO to bad business
  • Client Directors were particularly strong as the single point of responsibility for a client.  They usually had strong industry credentials, proven project management skills and enough technical skills to keep the projects correctly focused on the best possible business outcomes.
  • Team structures included financial director and multiple delivery leads
  • A P&L for every project! Fanatic about margins and profits! Monthly review!

 

EXTREME FOCUS ON BUSINESS VALUE for CLIENT – Not on Technology.  Example:  My contact’s boss was “no BS, hyper focused on extreme customer value, brutal and disciplined after 35 years at Accenture.”

 

Deliver Extreme High BUSINESS Value And Charge For It!

  • ESTIMATING disciplines with QA on estimate: 50% minimum markup plus overhead plus contingency
  • Enforced contribution margin commitments by Managing Directors and Partners!
  • Made the high cost of committing people to marginal/bad projects clear to everyone

 

Accenture is very good at fixed fee projects but my contact noted that the Cost+ or Hourly work segment matters.  It serves as a low risk pool of talent to keep people on the team, trained for higher value work and to elevate the best people to highest value work.

 

Projects over $X went to Deal Review Committee which included people from all over the country.  If big enough the project had monthly oversight by the deal review committee. This takes advantage of all the expertise within Accenture instead of just “blasting ahead on your own.”  E.g. “Sally in Indiana has done this before”.  Some additional comments from my sources:

 

  • “Accenture runs as a Meritocracy!  They don’t just say the words – they live it.” 

 

  • “The leadership, strategy and future focus are amazing. I saw them retool the entire company twice in 13 years.”

 

  • “Accenture does scale incredibly well if the client will sign up for the outcomes needed”

 

  • “Because Accenture holds partners accountable it can do big, complex, involved projects.  It can take risks other Big 4 firms cannot.  For example, Accenture is ten times the size of Deloitte’s tech practice.”

 

  • “Accenture is now building its own products such as chatbot agents to replace humans in software development, chatbots to replace insurance claims people, chatbots to replace auditors and chatbots to replace tax prep people.  In some cases this can save the labor of 40,000 people”

 

Where Problems Show Up:  My contacts were candid about when Accenture makes mistakes or gets in trouble.  

 

  • CLIENT WON’T PAY FOR EFFECTIVE PROGRAM / PROJECT MANAGEMENT.  Client PMs are never as good, never have the authority to execute and are usually compromised by internal politics or self-interest.  The client PM tries their best but money and time get spent and when the problems become undeniable it is too late for an Accenture PM to deliver the project as originally sold.  THIS IS THE  BIGGEST RECURRING FAILURE I POINT HAVE SEEN.”    (TOM:  This is a well known and documented problem.  I have seen it in 100+ situations.)  SOME SOLUTIONS:  
    • “We started burying the PM cost as a REQUIRED OVERHEAD item to prevent the client from pulling out the expense.”  (TOM:  I recently made a similar mistake.  I made it too easy for the sales people and a client to identify the contingency funds and pull them out.)
    • “We would confront clients with a “non-–performance“ letter. In one case we risked $150 million per year to correct the situation.”
  • “BIG CLIENTS GET THE BEST PEOPLE AND BEST RESULTS.  Second tier clients are where the problems usually show up.”  Many of us know of big consulting firm failures.  As we dig into WHY, we see these failures are rare with the really big clients who are willing to pay to do it right.  These clients get the best of Accenture’s people (and their people tend to be better as well.)   Sadly, other clients get less than the best – and are often unwilling to pay to do it right.  Many, many lessons are under this fundamental truth.
  • Other Solutions That Improve Client Outcomes: 
    • “CONTRACTUAL Governance, Cadence, Meetings, Change Control, Release Timeline, Issue, Resolution, Escalation, Earned Value, Work Packages, Work Breakdown Structure, Budgets…” 
    • “We FORCED client executives to review project reports and see costs of WHAT THEIR PEOPLE WERE DEMANDING.”
    • “We showed the percent of features Accenture delivered versus those delivered by in-house Project Management Organizations (PMO).”

Sometimes, It Still Goes Wrong:  We finished by discussing a BIG NIGHTMARE at a Fortune 500 Dallas client where Accenture REFUNDED MILLIONS OF DOLLARS.  Even the best of the best are not perfect but there is much to learn from Accenture.

 

 

627

Trap:  “Sure, We Can Do That…”

 

Trap:  Lack of Real Product

Long Term Manufacturer’s Sales Agent on REQUIRING REAL PRODUCT before He Will Take On A Client

 

In 2024 Trent H, a long time friend and 20 year+ veteran of selling in and around the semiconductor industry mentioned this issue.  Mos of us have seen that software people always respond with “yes, we can do that…”  I found it instructive that in an electronics / hardware based business like semiconductors he found the same problem.  As Trent put it “I can't sell 'one more engineer's great idea'.

 

 

626

Strategy

 

Requirements

 

Scope

 

Process

 

Sales / Marketing

 

Leadership

 

11/29/2025 WSJ:  Anduril (Defense Contractor) Drone Boat software failure results in public embarrassment, 30 drone boats towed out of exercise zone to prevent hazard to navigation

 

 

 

 

 

 

625

Cost / Margin

 

Strategy

 

Leadership

 

Barriers to Competition

 

Authority / Responsibility / Competence (OPERATING EXPERTISE REQUIRED)

 

Discipline of Profitability Day 1

10/9/2025  CISCO FOUNDER JOHN CHAMBERS:  VENTURE CAPITAL (VC) MODEL FAILING, TOO LONG, TOO MUCH RISK

 

  •           Predicts 1/3 of VC Will Fail, Implode.  (He is a VC now.)
  •           VC bad for four years, exits not happening, Limited Partners not happy
  •           IPOs take TOO LONG  (12-15 years), risk TOO HIGH
  •           VC successes have come from prior generation investments
  •           MANY TRAINWRECKS - come from OVERFUNDING, OVERSPENDING
  •           Survival requires OPERATIONAL EXPERTISE, A RAPID PATH TO PROFITS
  •           NOT TECH FOR TECH’S SAKE – looking for OUTCOMES
  •           B to B AI and Cyber Security are his focus.
  •           AI:  Where’s the PAYBACK / 10X SOLUTION?  Customer service and coding:   Should see 30-40% operating improvements per year for end companies
  •           Link to Article

 

 

624

Discipline of Profitability Day 1

September, 2025, Harvard Business Review,.   Why Startups Benefit When Big Investments Come Later.  Summary of research by HARSH KETKAR of UT Austin and Maria ROCHE.  CONSISTENT WITH

  • DISCIPLINE OF EARLY PROFITABILITY by Christensen and Moore.
  • GOOD MONEY / BAD MONEY concept by Christensen.
  • RIGHT STRATEGIES EVOLVE – RARELY RIGHT FROM THE START by Christensen

 

- Concludes startups are MORE SUCCESSFUL when funding stays low early in company lifecycle.

- Argues that experimentation, innovation, unconstrained are essential to evolution of effective strategy.

- Also notes that investor tolerance for experimentation, exit strategy, and degree of micromanagement are make or break issues.

- Study of 11,853 US tech companies from Pitch Book.  Note: most were creating apps and websites. Not seeing significant rigor in the research. Appears to have come to the RIGHT conclusions for wrong reasons. Study is overly focused on innovation as solution to all startup problems.

 

 

620

LANDMARK on  STRATEGY

 

Compete on Product / Tech / Features

vs.

Compete on Cost / Economies

vs.

Compete on Highest Value for Customer

1960 Landmark HBR Article - Theodore Levitt Problem / Opportunity:  MARKETING MYOPIA.  Mistake of focusing on YOUR PRODUCT instead of CUSTOMER NEED.

  •           Argued that companies fail when they focus too narrowly on their products instead of the customer’s underlying needs.
  •           Classic example: railroads thought they were in the "railroad business," not the "transportation business," which left them vulnerable to cars, trucks, and planes.

Also Popularized the Product Life Cycle (PLC):  Products go through predictable stages: introduction, growth, maturity, and decline.

 

 

617

AI Updates

 

Who Gets The Margin?

 

USE CASES:  Show Cost / Benefit

 

STRATEGY:  Compete on Highest Value for Customer

9/2/2025  Containing The AI Hype:  Big software company SN spent three years, big cost, big hype adding AI to product line.  Out of Thousands of customers, to date have only sold two “Agent AI” customers and a dozen paying for “Advisor AI”. Per MM, details confidential.  See HOW THE WINNERS DO IT, Other Useful Lessons

January, 2025  McKinsey Paper, Other Citations show PROBLEM IS FOCUSING ON TECH – RATHER PROBLEMS and SOLUTIONS CUSTOMERS ARE WILLING TO PAY FOR – AT HEALTHY MARGINS

Also known as the THEODORE LEVITT TRAP (a Harvard Professor):  When A Business thinks of itself as SELLING PRODUCTS rather than SATISFYING CUSTOMERS.  Click for Details

2/10/2025 WSJ, Andy Kessler:  The Best Use Cases:  Call Centers, Doctor’s Notes Summaries, Legal Tasks, Companions, Software Coders, Education, Graphic Designers

9/2/2025  Other Key Use Cases Identified in Last Three Months:  Detecting Deep Fake Video / Audio Forgeries, Checking Code for Hacking Vulnerability

 

 

616

How to Get Dramatically Better Outcomes from Big Application Software

 

Fix The Process First!!!

Fix The Process First:  Click for article in process, citations from Bill Davidow, Geoffery Moore, Clayton Christensen, Fred Wiersema, Tom Davenport, Lynne Markus

  • SAP (ERP software) discovered in 1992 that their customers who were already good at fixing the process first had dramatically better re sults from implementation.  SAP partnered with 14 large service providers to help the vast majority of customers who were not good at fixing processes first.  While financially successful for SAP and the services providers, customer results and feedback from service provider implementations have been mixed to negative.  (Crossing The Chasm, CITATION PENDING)
  • End Customers Who Are Already Good at Fixing The Process First:  At some point these companies have been noted for this skill.
    • General Electric, TE Connectivity, SAB Miller, Danaher (per David Rice), Cisco
    • What they have in common:  All identified so far have implemented “Lean”.  Most are acquirers who have learned over time that “fix the process first” is essential to integrating acquisitions effectively.
  • Process Improvement Practice in Seatle, WA:  (Name confidential) discovered several issues that helped significantly:
    • SEPARATE PROCESS IMPROVEMENT PROJECTS FROM TECH PROJECTS – do process first!
    • DOCUMENTATION + PROCESS:  Look for clients who are already pro-process and pro-documentation.  Lean / Six Sigma / Quality initiatives are best leading indicator
    • MINIMUM $50,000 TO $100,000 PROJECTS.  Qualifies and separates from those not really committed

612

FGIC Problem (Financial Guys in Charge)

 

Strategy

 

Leadership:  Authority / Responsibility / Competence Matching

 

Cost / Margin

 

Labor / Capital / Technology Cycle Arguments:  (article in process - see OneNote)

  •           Long-Wave Economic Theory (Kondratiev Waves):  (per Chat GPT)  Capitalist economies go through 40–60 year cycles of boom and bust.  Innovation & production → over-financialization → crash → renewed focus on real economy

 

  •           I think we are 40 or so years in – resulting in finance having too much power, reaping unearned rewards.  I hope we will soon return to PRODUCERS OF VALUE in charge.

 

  •           Technology Cycle Argument:  Life cycles of new technologies such as the printing press, telephone, auto, airplane, television, microprocessors, the internet, cell phones, etc. are widely recognized.  Andy Kessler of the Wall Street Journal and others argue that the microprocessor is nearing the end of its cycle.  (No doubt AI is the next big thing.)

 

  •           I hope to see these cycles return to VALUE FOR CUSTOMER, VALUE FOR PRODUCERS within my lifetime.

 

 

501

Big Lessons from Our Research

 

A screenshot of a computer

Description automatically generated

 

Our study of 427 PUBLIC AND PRIVATE SOFTWARE COMPANIES

2025 and Ongoing:  Our study of 427 PUBLIC AND PRIVATE SOFTWARE COMPANIES reveals hard truths

  • Software industry UNPROFITABLE since 2015.   Robert Morris Associates (RMA) has been publishing financial results studies of the software industry since 1989 (when software was a healthy industry).  In 2015 the industry profits turned negative (due to the switch to SAAS subscription billing), but have not recovered, with annual losses as high as 5% for the industry.  Don’t be distracted by the profits of Microsoft, Oracle, Salesforce and SAP.  Big profits by a few distort reality in an industry.  Your chances of picking a winner are less than 1%. 
  • Only 11% of Public Software Firms beat the RETURN ON CASH invested in a MUTUAL FUND over 10 years.  Our research shows 89% of software companies pursue the wrong strategy.  We study and discuss the strategies of Winners – how these 30 companies manage to provide a return greater that 7% consistently over many years.  Several exceed 20% per year.
  • Software industry debt has climbed from 60% of assets in 1989 to 95%+ today.   Intangibles have risen from 7% of assets in 1989 to 40% of assets today.  This is due to extreme overpayment for acquisitions (and other bad practices.)
  • WASTED BILLIONS OF OWNER CAPITAL among public software / tech companies is BEYOND BELIEF.
  • CAUTION, CAUTION, CAUTION.  I think making a solid profit on fundamental value is so hard most have given up.  All the attention goes to techno-hype, product feature competition, growth mania, hockey stick projections and – when the inevitable nonperformance occurs – trying to sell the company to the greater fool.
  • How the Winners Beat The Return on Cash Invested in a Mutual Fund:  
  • Why 89% of public software / tech companies are pursuing the WRONG STRATEGY and WHAT TO DO ABOUT IT

Click for Full Study  (password required)

 

629

 

Ongoing:  Berkshire Hathaway, Warren Buffett, Charlie Munger Lessons for the Software Business (Summary by Tom Ingram)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

506

Big Lessons from Management Literature

3/10/2025  The Innovator’s Solution by Clayton Christensen

 

 

 

 

 

 

 

 

507

Big Lessons from Management Literature

 

3/17/2025  Crossing the Chasm, 3rd Edition by Geoffrey Moore:

 

 

 

 

 

 

 

 

 

508

Big Lessons from Management Literature

3/24/2025  The Discipline of Market Leaders by Wiersema and Tracy:

 

 

 

 

 

 

 

 

 

 

 

499

Find Your Niche

 

Shorten Sales Cycles

 

Narrow Focus on Extreme High Value for Customer

 

10x-30x ROI for Customers

 

Barriers to Competition

Self Referring Niches

 

Urgent Compelling Need

 

Ideal Prospect

 

Solve Sales Problem

 

Sales Execution Excellence

 

Avoid Sales Traps

 

Whole Product Solution

Documentum Case:  From Sales of $2 Million to $75 Million and IPO in 4 Years!  Shortened Sales Cycles from 18 Months to 6 Months.  Customer Payback Average Of 10 To 1, As High As 30 To 1.  Saw Strategy Working Within 60 Days

Strategy

  • Narrowed 1 million+ prospects to 41 pharmaceutical manufacturers. Closed 39 in 1 ½ years.  NOTE:  Financial services was a BIGGER niche, but need ws NOT AS URGENT as pharmaceutical need. 
  • Culture focused on customer problems – not “what we do”
  • Focused entire company on single niche. 
  • Grew other niches through referrals from first niche.  Chemical manufacturers needing regulatory approval was the second niche
  • Value of defining a niche small enough that you can dominate it
  • UCN (Urgent, Compelling Need) defined as customer must act in 90 days
  • Ideal Prospect model to define, clarify target 

Product / Offering

  • Don’t go crazy on perfect pitch – go, learn, improve. 
  • Just get after the single niche – make calls – improve the offering rapidly
  • Reshape offering to get to a single line of business decision maker. 

 

Sales Execution

 

  • How Shorten sales cycles:  "Urgent, compelling problem that we solve" clear to sales, entire org. 
  • Medium sales person executing focus strategy superior to senior person with big Rolodex. 
  • Marketing’s Job:  Give Sales Person a List, Who Call on, What Listen For. 
  • TRAP:  Unfocused approach defeats the referral process
  • TRAP:  Taking big deals that pull you off focus.  Rule on taking business outside your top niche
  • TRAP:  Not doing the homework to give salespeople effective prospect list,
  • WHOLE PRODUCT SOLUTION:  Not just the software.  Everything the customer needs to say “I just give these people money and my problem goes away.” 
  • TRAP:  Failing to ruthlessly qualify. 
  • TRAP:  Failing to simplify message to three things that solve the problem. 

 

More information

 

 

(Source: Interviews and Crossing the Chasm)

 

497

Narrow Focus on Extreme High Value for Customer

 

Beating Bigger Competitors

 

Whole Product Solution

 

Bowling Alley of Adjacent Niches

 

Fix The Process First

 

Barriers to Competition

 

Seeing Big Picture, Macro Trend

 

Circle of Competence

 

Actively Fight Complexity

 

“Sign Posts” To Differentiate from Competitors

 

Winning 

The “Integration Devastation”

 

Single Line Of Business Buyer

 

Urgent Compelling Need

 

Rare Win in Volatile Industry

 

Sign Post Signal of Differentiation

Lawson Healthcare Niche Success:  Grew from $4 million to $40 million in four years.  Dominated Super-Narrow Niche Of Healthcare Integrated Delivery Networks (HIDNs).  Successfully Competed with Oracle and PeopleSoft

 

BRILLIANT NARROW FOCUS on BEACHHEAD segment of HIDNs.  This is a great example of whole product, creating barriers to competition and selling multiple bowling pins (products) to the same customer.

 

Fixed The Process First With Activity Based Costing Training / Consulting As Part Of The Whole Product Solution. 

 

Spectacular Example Of No Code Workflow Software + Costing Helping Customer To Reengineer Processes.  No code software was kept simple, did not require coding skills of most competitors.  (Tom:  I have seen this problem many, many times).  Other notes:

 

  • Training in Activity Based Costing for users was a key differentiator and barrier to competition.
  • Lawson retained market leadership even after a fast follower attack by PeopleSoft.
  • SUCCESSFUL NAVIGATION OF HEALTHCARE LANDMINE INDUSTRY WHILE WARREN BUFFETT AND TOM  STAYED AWAY.  Saw healthcare systems in turmoil in the 90s. Computer systems needing dramatic overhaul.
  • Treated smallness as an advantage, forcing it to focus.  FOCUS IS VERY HARD IN LARGE INSTITUTIONS always looking for big returns right out of the gate.
  • SINGLE line of business buyer: CFO and staff
  • Urgent Compelling Need: Customers were desperate to get pricing under control with the new business model. 
  • Minimum set of products and services needed to fulfill the UCN
  • Activity Based Costing module was essential for customer to understand revenue and cost by patient, procedure, fixed asset, health plan. Had to adapt to new pay model going forward.
  • BIG WIN FOR NO CODE SOLUTIONS PLUS PROCESS REENGINEERING
  • Included Materials Management software for unique needs of health care – e.g. a Surgery Cart.  BECAME SIGN POST SIGNAL of differentiation
  • INTERFACES to legacy application systems: Most critical was patient management. Became so strong in the segment that major legacy applications allied with Lawson over Peoplesoft.
  • Whole Product Solution Nuanced Win:  All competitors had standard training.  Added training on Activity Based Costing to round out whole product and create barriers to competition. 

(Source:  Crossing the Chasm, Innovator’s Solution) 

 

496

Profitable From Start

 

Overcame Technology for Technology’s Sake 

 

Actively Fight Complexity

 

Costs / Margin 

 

Practical Beats “Cool”

Brickstream breaks cycle of failure and finds the right new strategy. 

 

 

  • Failed with initial vaporware, techy, bleeding edge video solution for retailers 
  • New management refocused on solving a much simpler problem 
  • New solution was not as compelling or dramatic but it was a VIABLE BUSINESS 
  • Key was DEMANDED PROFITABILITY VERY EARLY ON to validate 

 

(Source:  Crossing the Chasm) 

 

495

Narrow Focus on Extreme High Value for Customer

 

Urgent Compelling Need

 

Bowling Alley of Self Referring Niches

 

Pragmatic Buyer – Not Early Adopter

 

Effective Reseller Programs, Contains Problems

 

 

 

 

VMWARE Examples:  220% ROI in 9 Years.  New Product $0 to $40 million sold in Texas alone in two years. 1,000 New Customers

 

 

  • Had a string of niche product wins before Crossing Chasm 
  • Urgent Compelling Need that built the company was testing and other short term needs to spin up, spin down servers quickly and cheaply  
  • Second bowling pin niche was unutilized capacity 
  • Ask us about Bruce Switzer's win with Viper product (software based storage). $0 to $40 million in sales in 18 months in Texas alone.  
  • Has 3-4 failures for every success like this.
  • Clear, high, pragmatic buyer value. 
  • Took time - year or two - to organize and execute.
  • Strong Profits = High, Stable Market Value
  • 2/26/2024 recently acquired by Broadcom
  • Feb 2024, WSJ:   BUNDLING:  Consolidated 160 products to TWO primary bundles.  ADDING FEATURES AND CUTTING PRICE – AS PORTER PREDICTED
  • 3/21/2024 WSJ:  has done a GREAT job of productizing complex software so can be effectively sold through distribution.
  •           Keeps reseller products and services simple.  Retains high complexity, high margins services for itself improving customer results and margins.

 

 

(Source:  Crossing the Chasm, Innovator’s Solution, Interview with former executive) 

 

494

Profitable from the Start

Absolute Software:  GREAT job of profitability from the start

 

See TIA publication How to Increase Margins for Software Companies 

 

(Source:  Crossing the Chasm) 

 

493

Whole Product Solution

 

Narrow Focus on Extreme High Value for Customer

 

Architecture, Product Pipeline

Corelation Inc (Credit Union swr, GREAT EXAMPLE OF Focus, WHOLE PRODUCT PARTNERS. 

 

CORE Processing Solution is designed to enhance credit unions and includes no code, dashboards, user engaging interface.  

 

Collaborates with strategic partners to integrate additional functionalities into their platform such as online and mobile banking, digital account opening, loan origination, and digital card issuing. 

 

Jack Prim IS A FAN:  Private, solid, great profits, where Jack Henry was 20 years ago.  Does not want to go public or raise capital "leave us alone and let us run this business") 

 

(Source:  Crossing the Chasm, Jack Prim, former CEO of Jack Henry) 

 

492

Whole Product Solution

 

Narrow Focus on Extreme High Value for Customer

 

Niche With No Dominant Competitor Which You Can Dominate

Intel Won The Microprocessor War With A Whole Product Solution. 

 

Product included the chip, application notes, ads, microprocessor development systems, emulators, software, field applications engineering, single board computers, customer education programs and marketing's capture of public imagination.

 

Intel beat HP and Tektronix in Market for a new segment of electronic instrumentation. Competitors built general purpose equipment supporting any manufacturer’s microprocessor. 

 

Intel focused narrowly on being spectacularly good at the whole product surrounding Intel microprocessors.  Competitors could not match Intel's whole product solutions.

 

Initially found niches with no dominant competitor.

 

(Source:  Marketing High Technology) 

 

491

Business Expertise

 

Narrow Focus on Extreme High Value for Customer

 

Cost / Margins

 

Barriers To Competition

 

 

No Tech for Tech’s Sake

 

Extreme Focus on Service to Customer

 

Whole Product Solution

 

Pragmatic Buyer Market instead of Early Adopter

 

Big Lessons from IBM

 

IBM salespeople were consultants. Their personal contribution could overcome price and performance deficiencies. 

 

SPECTACULAR DOMINANCE OF MAINFRAME INDUSTRY page 151.  70% market share, 95% of industry profits.  Proprietary products, strong cost advantages, high barriers to competition, 

 

DIFFERENTIATION.  ACCOMPLISHED WITH NARROW FOCUS, EXTREME HIGH VALUE FOR CUSTOMERS.  Primary focus on service, value to customer over technology.   

REMEMBER – THIS SERVICE DOMINANCE CAME DURING ERA OF EMERGING, UNRELIABLE MAINFRAME COMPUTERS.  Customers needed high level of service and were willing to pay 

 

REMEMBER - IBM’s System 3, 34, 36, 38, AS/400 Series.  Likely best, longest whole product application value for customers in history.  Certainly for mid-range.  PROSPERED while Wang, Apollo, Prime, Convex, Data General, DEC went away!  

 

Amusing how derided as “old tech” for decades, but continues to sell and provide value to customers today.  WHOLE PRODUCT IS CLEARLY SUPERIOR FOR THE PRAGMATIC BUYER, even though the generic product itself is old tech. 

 

Great Precedent: IBM AS/400 and HP 3000 thrive while Wang, Prime, Apollo, Convex, DG, DEC gone

 

IBM’s System 3, 34, 36, 38, AS/400 Series

  •           Mid-range business applications
  •           Best Customer Intimacy Case in All IT History
  •           Likely best, longest whole product application value for customers in history.  Certainly for mid-range.

 

Additional Positives: 

  1. Probably BEST EVER AT EXTREME HIGH VALUE FOR NARROW SET OF CUSTOMERS 
  2. Rock of Gibraltar reputation 
  3. Technology follower - not a leader. Became a strength 
  4. Best total solution for pragmatic, main street buyer
  5. Easily weathered technology changes 

 

Some Negatives 

  1. Nonproprietary movement was too big of a change, IBM was too slow (note that this is very common in corporate history) 
  2. (Arguably) took advantage of their monopoly power 

 

(Sources:  Crossing the Chasm, Discipline of Market Leaders, Marketing High Technology, Innovator’s Solution, Tom’s experience competing against IBM) 

 

490

Profitability from the Start

 

Narrow Focus on Extreme High Value for Customer

 

Big Lessons from GE

 

GE’s famous “#1 or #2 or get out”

 

(Source:  Marketing High Technology) 

 

489

Profitability from the Start

 

Narrow Focus on Extreme High Value for Customer

 

Saying NO, Not Chasing Every Dollar, Extra-Large Customers

 

Margin / Cost:  Can Afford to Invest In Service

 

Charges Significantly More Than Competitors

 

Urgent Compelling Need

 

Understands Customer Problems Better Than They Do

 

Are You Sufficiently Focused:  How To Know

 

Rational Acquisitions 

 

Ruthlessly Qualify Prospects

 

Sell the Business Executives, Not IT Dept

 

Service as Primary Barrier to Competition

 

Avoiding Wall Street, IPO Traps

Jack Henry:  500% Return on Invested Capital from 1995 to 2005.  300% Return on Invested Capital from 2012 to 2021. 

 

See our case study for these key lessons: 

 

  1. Focused on medium/smaller banks and credit unions where needed and valued high level of service. 

 

  1. AVOIDED BIG BANKS who think know it all, can demand anything

 

  1. Extreme High Service for customers who will pay for it.

 

  1. Healthy margins as leading indicator you are doing the right things.

 

  1. Watch for shenanigans by incumbent vendors.  Walk away.

 

Disciplines

 

  1.       Saying "No" to Big Customers

 

  1.       Saying "No" to Overpriced Acquisitions

 

  1.       Hard Data Shows the Customers Love What They Get for the Price

 

  1.       Sales Execution
    1.       Business Unit Executive Sell – NOT SELLING TO IT
  2.       Technical / Operations Execution
    1.       Customer Service – strongest possible

 

(Source:  Case Study:  No27JackHenryProfitable30Years.pdf , Interview with CEO Jack Prim, 2nd Interview with Jack Henry CEO Jack Prim)

 

488

Narrow Focus on Extreme High Value for Customer

 

Economies of Focus

 

Avoided “Tech for Tech’s Sake” Trap by Focus on Customer Need

 

Generic Products Not Viable

Reseller of No Code / Low Code Software and Services.  Grew from $1000 investment to $28 Million in Sales by Narrowing Focus  

 

See our case study for these key lessons: 

 

  1. Altered focus from generic product to narrow, discrete target markets. 
  2. The horizontal sales approach no longer works
  3. Narrowed universe of prospects from one million to 200. 
  4. Changed customers!
  5. Tight focus and producing super-high customer value, solving urgent needs
  6. Customers skeptical, tired of unmet promises
  7. Becoming and expert in a very narrow field
  8. Trap:  Focused on “what we sell” rather than “what customer needs”

 

(Source:  TIA Case Study  No23SwrFirmStartedwith1000.pdf)

487

Narrow Focus on Extreme High Value for Customer

 

Leadership Replaced

 

Tool Kit Products Not Viable

 

Shorten Sales Cycles

 

Barriers to Competition

 

Whole Product Solution

 

 

European Software Company Struggling in U.S.:   Closed Six Large Sales in Eight Months (After Closing Zero Sales in the Previous 16 Months).  

 

See our case study for these key lessons: 

 

  1. Discovered (the hard way) that Europeans would buy "Toolkit" Products while North American buyers demand Whole Product Solutions. 

 

  1. Most of Management Team Replaced After 16 Months of no sales

 

  1. Lead sales person came from target customer’s industry

 

  1. Dramatic shortening of sales cycles through narrow industry focus and Whole Product Solution

 

  1. Consultative Selling worked (defined as focus on customer outcomes and results as first priority.)

 

  1. Whole Product Solution thinking resulted in offering that was substantially better than any competitor

 

  1. Found Right Resellers.  Resellers significantly contributed to the Whole Product Solution

 

  1. Hard part was getting management to listen to those close to the customer

 

  1. Shortened sales cycles by adding six additional swr modules to create a WHOLE PRODUCT SOLUTION.  

 

  1. Customer tells themself "I just give them money and my problem goes away."

 

(Source:  TIA Case Study  No24SwrCoCloses6LargeSalesin8mos.pdf)

 

486

Narrow Focus on Extreme High Value for Customer

 

Shorten Sales Cycles

 

10X Payback for Customer

 

Ideal Prospect Concept

 

Direct Approach Worked

 

Distribution Channel Through Consultants

 

Did Not Chase Every Dollar of Revenue – Selective on Customization Requests

 

 

 

Struggling Software Firm Gets Focused, Sells 12 New Accounts in 16 Months, Generates Over $10 Million.  

 

See our case study for these key lessons: 

 

  1. Product was SAP add-on for change management.  Solved very large problem of user change requests.
  2. Key was focus on one strong product sold to a tightly defined niche. 
  3. Successfully targeted IT Buyer (where many have failed). 
  4. Closed every pilot project sold except one
  5. Huge ROI on Solution.  Saved customers between 13% and 20% of labor spent on changes.  (Easily saved $1 million to $10 million per year.)
  6. Ideal prospect: WOULD ADMIT THEY HAD A PROBLEM, could afford $100,000 cost. 
  7. Direct contact approach worked
  8. Built distribution channel through big four consultants.  Consulting firm earned big implementation revenues and endorsed the product.
  9. Picked the gem among all the acquisitions.  Ignored the rest.
  10. Had capital, was able to define product from customer requests.  Selectively accepted customization requests if they strengthen the core product

 

(Source:  TIA Case Study No26StrugglingFirmSells12NewAcctsin16Mos.pdf)

485

Urgent Compelling Need

 

Narrow Focus on Extreme High Value for Customer

 

No Price Competition

 

Whole Product Solution

 

Barriers To Competition

 

Bowling Alley of Self Referring Niches

 

Leadership

 

Say “No” To Business Outside Your Focus

 

Shortened Sales Cycles

 

Minimal Acquisitions, Done Well

Docucorp:  Struggled for Five years, Got Focused, Grew Sales from $15 Million to $75 Million During Difficult Post-Tech-Bust Years! 

 

From initial losses, rose to solid profits, spectacular return on low invested capital

 

See our case study for these key lessons: 

  1. Hyper-focus on nothing but policy production for insurance companies, Dominated Niche. 
  2. Narrowed the Universe of Prospects from 1 Million to 500.  
  3. Narrow focus creates efficiencies, dominates competitors at low cost.
  4. Found an urgent need to buy niche that did not require major changes to their product
  5. No Price Competition
  6. Moved away from “product-feature” selling to “what is the whole product solution the customer needs to solve their problem?”
  7. Picked the Right Niche, Dominated It, Then Moved On to Subsequent Niches
  8. Narrowed the Universe of Prospects from 1 Million to Less than 500
  9. Stopped Trying to Sell Technology to Anybody Who Would Listen
  10. Picked the Right Niche, Dominated It, Then Moved On to Subsequent Self-Referring Niches
  11. Focused on Profitability and Cash Flow, Simple, Straightforward Business Plan
  12. Six to Twelve Months to Know They Were on the Right Course
  13. Had to Invest in Developing the Leadership and Management Team
  14. Walked away from business outside their focus
  15. Shortened Sales Cycles by focusing on an industry, stopped selling to IT, found Urgent Compelling Needs
  16. Few acquisitions, done well, while competitors floundered with numerous bad deals

 

 

(Source TIA Case Study:  No28DocucorpSuccessStory.pdf )

484

 

10X Payback for Customer

 

Narrow Focus on Extreme High Value for Customer

 

Honest Sales Representations

 

Solving the Sales Problem

 

Minimal Price Competition

Fiserv:  11.14% Average Net Return on Sales during DOT COM bust.  Grew from $20 Million to $3.5 Billion in Sales, Profitable Every Quarter.  Share Price Up from $1.22 to $37.00 a Share after Seven Stock Splits

 

Spectacular market value on moderate invested capital.  Sustained market value as Tech Bubble burst.

 

See our case study interview with the CEO for these key lessons: 

 

  1.       “We are Not a software company.  We are a firm that produces results for our customers and we happen to use software”
  2.       “We serve a very narrow customer set.”
  3.       “Before you focus on revenues or earnings, you focus and quality and customer satisfaction”
  4.       “We solve inherent problems with internal IT departments”
  5.       The Business Process Reengineering Problem: 
    1.   “We help the customer when industry consolidation creates FAR MORE WORK THAN THEIR PEOPLE CAN DO”
    2.   “We help them meet the need for rapid change and ever reducing costs”
  6.       “We do acquisitions right”
  7.       “We provide executive bonuses and recognition tied to customer satisfaction and customer retention”
  8.       “All promises and representations made to the customer are done so in writing and in a fashion that is enforceable, testable and verifiable.”
  9.       KEY LESSON:  Where can you find a niche, like FiServ, where you can solve big and important problems for your customers and they will be delighted to pay you for it?
  10.   We have strong sales and marketing keeping the pipeline full because there are things beyond our control...”

 

Fiserv provides services and software to financial services industry.

 

(Source TIA Case Study:  No29FiServSuccessStory.pdf   NOTE:  This interview took place at Fiserv’s height in 2004.  Fiserv has come under some recent criticism…  We can still learn from what built this spectacular success.)

483

Early Profitability

 

Narrow Focus on Extreme High Value for Customer

 

Barriers To Competition

 

Solved Tech for Tech’s Sake Problem

 

Sales Funding, Execution

 

Whole Product Services Critical to Success

Ixos – Imaging Software.  20+ software sales of $1 Million+ at 50%+ Gross Profit.  Profitable after 3 ½ years.  

 

Workflow Add-On for SAP AP, AR, Other Financial Apps.  

 

EXTREME SUCCESS, LESSONS BENCHMARK versus competitors with technology focused strategies. 

Prospered with Focused Application Value niche while competitors (Viewstar, IBM) floundered.  

 

Filenet, a competitor, also prospered with Focused Application Value niche while competitors (Viewstar, IBM) floundered

Initially 5-10% services, grew to 40-50%

 

Sales Execution

 

 

(Source:  Interviews, Tom worked for competitor)

482

Narrow Focus on Extreme High Value for Customer

 

Bowling Alley of Self Referring Niches

 

No Tech for Tech’s Sake

 

Said “No”

 

Profitable From Start

 

Tyler Technologies:  TIA study winner, 100% ROI in 5 Years

200% ROI in 9 Years. 

 

Narrow focus on city government, started with city tax management and electric utilities.  Grew to eight city management segments total. 

 

  • KEY TO SUCCESS:  FOCUSED ON END CUSTOMER EXPECTATIONS - NOT GENERIC CUSTOMER SATISFACTION
  • AVOIDED TECHNO FASCINATION TRAPS THORUGH FOCUS ON CUSTOMER
  • Steady Profits = Big, Stable Market Value
  • Said No on deals that pull off focus
  • Lots of growth by smart acquisition
  • Stayed FOCUSED on core capabilities
  • Well attuned to the voice of the END USERS and citizens
  • Municipal govts 10,000+ universe of prospects

 

  • PRICE - NOT NOTABLY HIGHER than competitors

 

  • DID NOT TARGET SINGLE LINE OF BUSINESS DECISION MAKER.  Were able to sell to bureaucracy and committees in municipal govts, responsive to all inputs 

(Source: Customer, Interview with Marketing Consultant)

 

481

Profitable from the Start

 

Narrow Focus on Extreme High Value for Customer

 

BENTLEY SYSTEMS INC (PROJECT MGMT FOR ENGINEERS AND CONSTRUCTION)    2023 TIA Research Winner.  100% ROI in two years. 

 

Focus on STRUCTURAL AND CIVIL ENGINEERS USING SOFTWARE TO SIMULATE BIG PROJECTS.  

 

4/19/2024 WSJ:  Schnieder Electric bidding for control where it will merge its software business with Bentley.  Family controlled Bentley may not go for it.  APPEARS SCHNIEDER UNDERSTANDS BENTLEY IS WAY BETTER AT SOFTWARE THAN IT IS.

 

(Source:  TIA research)

480

Narrow Focus on Extreme High Value for Customer

 

Urgent Compelling Need

 

Whole Product Solution

 

Do Not Chase Every Dollar of Revenue

 

Saying “No”

 

Scope / Change Control

 

Big Winner Without Most Current Tech

Black Knight, Inc.  (LOAN ADMINISTRATION SWR)    2023 TIA Research Winner.  100% ROI in four years.  

 

Urgent Compelling Need:  Regulators compelled Bank of America to move off their home grown mortgage software and implement a package approved by the regulators.  

 

BofA management wanted one reliable neck to hold responsible.  

 

Fiserv was a distant second choice.  

 

The WHOLE PRODUCT SOLUTION included regulatory compliance and a rigid implementation process.  

 

Black Knight used a super rigid implementation and NEVER FAILED to get a customer implemented.  It was a 

genuine, strong process but not sophisticated.  “Their way or the highway”.  

 

Said “NO” to customizations.  Significant hard feelings with very big banks who want it their way, think they are smarter than everyone else

 

Stodgy, old tech.  Mainframe app with new interface

Now adding some bolt on apps for more value

 

(Source:  Bank of America former employee) 

 

479

Narrow Focus on Extreme High Value for Customer

 

CADENCE DESIGN SYSTEMS INC  2023 TIA Research Winner.  100% ROI in two years. 

 

Focus is electronic engineering design

 

- STRONGEST PROFITS IN STUDY

- Low invested capital

- Huge valuation

 

(Source:  TIA Research)

 

478

Narrow Focus on Extreme High Value for Customer

 

 

Cerner  2023 TIA Research Winner.  100% ROI in six years

 

  •           Became a dominant company in medical records, medical systems
  •           Winner from original study
  •           Large invested capital
  •           Net Profit:  13% avg at update, originally very strong profits, avg 18%
  •           Gross Profit: 83.5% 
  •           Software/Services Mix:  56% services
  •           SG&A 40% of revenues avg
  •           Strategy:  Software Sells High Margin Services
  •           Acquired by Oracle – going downhill fast

 

(Source:  TIA research, met founder, interview ex-employee)

 

477

Fix The Process First

 

Narrow Focus on Extreme High Value for Customer

 

DESCARTES SYSTEMS GROUP INC  (Canada, 2023 TIA Research Winner.  100% ROI in eight years 

 

BUSINESS PROCESS AUTOMATION SWR, supply chain  management business processes) 

 

(Source:  TIA research)

 

476

Narrow Focus on Extreme High Value for Customer

 

GLOBALSCAPE INC   2023 TIA Research Winner.  100% ROI in 6 Years.  300% ROI in 9 Years 

 

CRM software, provides secure information exchange, data transfer and sharing capabilities

 

Successfully acquired due to STRONG PROFITs and LOW INVESTED CAPITAL

 

(Source:  TIA research)

 

475

Profitable from the Start

 

Narrow Focus on Extreme High Value for Customer

 

Barriers To Competition

 

Group 1 Software 2023 TIA Research Winner.  100% ROI in five years. 

 

Direct marketing and customer relationship systems

 

ACCOMPLISHED WHILE COMPETING WITH SALESFORCE! 

Winner from original study

  •           Low invested capital
  •           Acquired
  •           Big Win:   spectacular market value from very low invested capital
  •           12%-15%, one down year,
  •           Modest acquisitions

 

(Source: TIA research)

 

474

Profitable from the Start

 

Narrow Focus on Extreme High Value for Customer

 

Barriers To Competition

 

MANHATTAN ASSOCIATES INC  2023 TIA Research Winner.  100% ROI in 3 Years, 200% ROI in 9 Years. 

 

SUPPLY CHAIN management software, solutions to manage supply chains, inventory retailers, wholesalers, manufacturers

 

FOUND NEW MARKET HELPING MANUFACTURERS SELL DIRECT TO CONSUMER. 

 

ACCOMPLISHED WHILE COMPETING WITH SAP AND ORACLE! 

 

Big Win:  spectacular market value on low invested capital, held value through tech bubble burst

 

Profits:  30%+ before tech bubble, 10% after, significant acquisitions

 

 

(Source: TIA research)

 

473

Narrow Focus on Extreme High Value for Customer

 

NATIONAL INSTRUMENTS CORP   2023 TIA Research Winner.  100% ROI in 9 Years 

 

ENGINEERING software for automated test equipment and virtual instrumentation software 

 

 

(Source: TIA research)

 

472

Narrow Focus on Extreme High Value for Customer

 

Barriers To Competition

 

Profitable from Start

 

Paycom Software, Inc.  2023 TIA Research Winner.  100% ROI in 6 Years

 

Software puts all HR functions online including talent acquisition, time and labor management, payroll, talent management and human resources management

 

Accomplished while competing with Workday, Peoplesoft, Oracle!

 

(Source: TIA research)

 

471

Narrow Focus on Extreme High Value for Customer

 

PTC INC.  2023 TIA Research Winner.  100% ROI in 9 Years 

 

MANUFACTURING software for computer-aided design / engineering, manufacturing, and services

 

Steady Profits = Durable Market Value

 

(Source: TIA research)

 

470

Narrow Focus on Extreme High Value for Customer

 

SHOPIFY INC.  2023 TIA Research Winner.  100% ROI in 9 Years. 

 

MARKETING TECHNOLOGY software enables merchants to sell their products across different sales channels 

 

Found niche HELPING MFRS SELL DIRECT TO CONSUMER

 

(Source: TIA research)

 

469

Profitable from Start

 

Narrow Focus on Extreme High Value for Customer

 

Bowling Alley of Self Referring Niches

VEEVA SYSTEMS INC  2023 TIA Research Winner.  100% ROI in 4 years.  140% ROI in 9 years

 

PHARMACY software cloud solutions for the life sciences industry / pharmaceutical, biotechnology, and medical device companies

 

Very Strong Profits created Very High, Stable Market Value

 

(Source: TIA research)

 

468

Narrow Focus on Extreme High Value for Customer

 

Dominate Niche

Boston Consulting Group study 1968:  MUST pursue sufficiently isolated segment which can be dominated. 

 

If cannot be dominated – withdraw.  

 

(Source:  Marketing High Technology, page 14) 

 

467

Narrow Focus on Extreme High Value for Customer

 

Ideal Prospect

 

Barriers to Competition

Reynolds And Reynolds - Big Win:  Spectacular market value on very low invested capital due to debt pay down and stock repurchase. 

 

SEE CHARTS, HUGE WIN FOR INVESTORS, Sustained Market Value after Tech Bubble!

 

Focused On Auto Dealers Who Needed Help And Were Willing To Pay 

 

Sales Execution

  • Tom:  I called on a car dealer as a competitor - R&R so strong I never called on another car dealer in 40 years.

Technical / Operations Execution

 

Big gains early, but big drop due to divestiture, regained some revenue growth

 

See study PROFITS OF 427  PUBLIC AND PRIVATE SOFTWARE COMPANIES, 2022

 

(Source:  Innovator’s Solution, TIA Research) 

 

466

Narrow Focus on Extreme High Value for Customer

 

Pragmatic Buyer Instead of Early Adopter

Oracle: Standardized on IBM’S SQL interface. 

 

Ported To Every Relevant Operating System / Platform 

 

APPS DEVELOPED IN ONE EASILY MIGRATED TO ANOTHER 

 

KILLER USE CASE!  Won pragmatists in IT depts 

 

This is what built Oracle

 

SPECTACULAR WIN FOR NON-PROPRIETARY

 

Oracle, disciplined enough to say NO to bad integration projects, demands from big customers

Jim C., Real Page former COO story, mad at first, then respected their discipline and focus

 

 

(Source:  Crossing the Chasm, Jim C, interview) 

 

465.5

Narrow Focus on Extreme High Value for Customer

 

Profitable from Start

 

SYNOPSYS INC  SEMICONDUCTOR TOOLS SWR   2023 TIA Research Winner  100% ROI in 8 Years

 

Semiconductor Engineering Simulations Design

 

Focused Application Value

 

Steady Profits = Big, Stable Market Value

 

12/23/2023 WSJ update:  Market value up from $60bb to $80bb, still going strong, chip design focus, acquiring Ansys

AMAZINGLY SMOOTH MARKET VALUE GROWTH FROM STEADY PROFITS.  VERY MUCH WORTH STUDYING

 

Source:  TIA Research 2023

465

Narrow Focus on Extreme High Value for Customer

 

Profitable from Start

 

ANSYS INC  2023 TIA Research Winner.  - STRONGEST PROFITS IN STUDY.  100% ROI in eight years. 

 

Focused on Manufacturing (especially simulations for Aerospace, Healthcare, Automotive.) 

 

MANUFACTURING SWR, engineering simulation software, simulates how real products work

 

Increasingly important for structural, mechanical, power? reasons as products smaller and more complex.

Very large invested capital

12/23/2023  Synopsis acquiring, TWO STRONG PERFORMERS, BUFFETT / MUNGER STYLE – ACQUIRE STRONG COMPANY AT FAIR PRICE?

 

(Source:  TIA Research) 

 

464

Narrow Focus on Extreme High Value for Customer

 

APPFOLIO INC  2023 TIA Research Winner  - 150% ROI in 10 Years 

 

PROPERTY MANAGEMENT SWR / Real Estate Swr

 

(Source:  TIA Research)

463

Narrow Focus on Extreme High Value for Customer

 

Cheetah Mobile Inc.  2023 TIA Research Winner.  100% ROI in five  years.  PROFITS GOOD BUT UNEVEN

 

Software applications for smartphones and tablet devices

 

Moderate invested capital, Market Value Dropped by $20 BB, Still Profitable, Viable.  

 

(Source:  TIA Research)

462

Narrow Focus on Extreme High Value for Customer

 

Bowling Alley of Self Referring Niches

NICE Ltd. (Israel)  2023 TIA Research Winner.  100% ROI in 9 Years 

 

CRM Swr  Customer Engagement and Financial Crime and Compliance markets   

 

(Source:  TIA Research)

 

461

Narrow Focus on Extreme High Value for Customer

 

Saying “No”

 

Go Find Good Business

 

35%+ Of Revenues on Sales / Marketing

 

Bowling Alley of Self-Referring Niches

 

EDS – Numerous Lessons

 

e.g. HOW BUILT $100 MILLION BUSINESS UNIT.  Services Only to Defense Contractors:  Engineering Manufacturing Development, Virtual Assembly Aviation Defense Contractor Niche

 

Initial Big Successes from Narrow, High Value Initial Offerings

  1. In 1962, understood the value providing, had courage to charge for it.  Sold fixed price five year contracts for services while competitors sold 60 to 90 day contracts.
  2. DISCIPLINE made the company.  No alcohol during business hours, conservative dress, loyalty, duty, near military management style created results and employee devotion.

THE DISCIPLINE OF NARROW FOCUS ON HIGH VALUE APPLICATIONS almost certainly proceeded from the core disciplines.

Core IBM principals (Don’t take bad business, Go find good business) also contributed.

  1. Early Customers, Collins Radio and Frito Lay helped but found real niches in:

Insurance data processing niche

Medicare processing started in 1965.  40% of business by 1977 and largest in nation by 1990.  Cumbersome 1965 legislation created opportunity for EDS

  1. First bank customer in 1968, grew to world’s largest data processing service for banks and S&Ls.
  2. Some failures, niches did not work
    1. Stock Brokerage
  3. Additional customer niches
  • Turnkey for hospitals
  • Turnkey for S&Ls 
  • Saudi Arabia
  • Iran – famously hired special forces guy to get his people out.
  • U.S. Army
    1. Beginning of the end:  Sale to GM, Switching to Integration Projects, Sale to HP

 

(Source:  Former employee interviews) 

 

 

Horizontal / Vertical Focus Done Well

 

Narrow Focus on Extreme High Value for Customer

 

Bowling Alley of Self Referring Niches

 

High Margin and Worth It

Peoplesoft:  Great Example Of Horizontal Platform Marketed Vertically To Industries

 

Captured 50% Market Share For Client / Server HR Apps, Then Moved On To Related Niches

 

- Lagged behind until Windows 3.0 established as standard for client of client server in early 1990s

 

- One of 3 core apps that Drove Windows 3.0 adoption (also Sap and Oracle)

 

- Able to keep charging premium because of market dominance

 

- HR systems were low risk app for pragmatists to try client / server

 

- Moved to next pin, financials, very strong due to HR position

 

- Challenge will be narrowing financials app targets to ones where HR strength is enough advantage to compete with Oracle and SAP

From Crossing the Chasm 3rd Edition pg 173-174 and Inside the Tornado (subsequent Geoff Miller book)

460

Narrow Focus on Extreme High Value for Customer

 

Workday (See Peoplesoft as Starting Point):  

 

Peoplesoft success was due to bringing big suite of interactive HR tools to client / server (which HR had never had). 

 

After hostile takeover by Oracle, former executives formed Workday and did the same thing with SAAS / online HR software.  

 

BIG HORIZONTAL WIN BY FOCUSING ON JOBS CUSTOMER NEEDED DONE (rather than industry vertical) 

 

- Key SAAS benefits

  * IMMEDIATE implementation

  * Pay as you go, lower risk

  * Continuous releases / updates

  * PERCIEVED LOWER SWITCHING COSTS (process is still the bigger barrier)

 

(Source:  Innovator’s Solution) 

 

459

Narrow Focus on Extreme High Value for Customer

 

Bowling Alley of Self Referring Niches

 

Urgent Compelling Need

Clarify:  Possible big win with textbook Chasm Crossing approach 

 

Software Application Support:  Full SAAS automation of Customer Service Software and Network Hardware Support.  

 

Won Cisco, 3Com, Synoptics, Wellfleet, Microsoft. 

 

NOTE:  Jury is still out – not sure if successful yet, but good lessons

 

Next Niche:  Telecom.  Same customer service / application goals as TE Connectivity 

- Agent Screen:  Customer Calling, (all possible info), Product causing call, Knowledge to meet need 

- Increase Sales through Better Service  

- Increase Sales with Buying Patterns, Promotions tuned to customer 

- Order / Shipping knowledge, troubleshooting 

- Screen sharing with customer 

*** Bug Case routing to development org, tracked through resolution 

- Skill based routing to best fit agent 

- Hot Lead capture, prioritization 

- Email, voice or chat as customer prefers 

- Automated chatbot service where works 

- Quotes tracked through on time response, esp. outside service dept. 

- Service Level Agreements enforced for resolution 

- Warm handoff of leads to partners 

- Agent productivity, measurement, training calls 

- Tiers of service, charge for, effective 

- Universal Agent:  All able to do customer service, product specialist, inside sales jobs, best possible 

- Remote agents on PCs or Tablets  

- Integration to tracking sensors, other customer systems 

- Speech to text, natural language, text analytics 

 

Urgent Compelling Need:  Competitive advantage, increased sales from superior service

OPPORTUNITIES:  

- Network Hwr Support, Vertical market with broken mission-critical process

- Niche lends to whole segment of customers – not just one at a time

LESSONS:  

- Make a total commitment to your focus niche, use whatever you have left to service other easy business coming in

- Started with Network Hwr Support, won Cisco, 3Com, Synoptics, Wellfleet

- Next bowling pin:  Software App Support, won Microsoft

- Next Pin:  Telecom

 

(Source:  Innovator’s Solution, Tom client)

 

458

Narrow Focus on Extreme High Value for Customer

 

Failed Horizontal Approach

Lotus Notes:  Big initial success with narrow focus on global account management for worldwide accounting and consulting firms. 

 

Migrated to another pin – customer service for high-tech companies.  Other customers started adopting, became fragmented, used for everything, lost focus, 

 

BECAME HORIZONTAL, acquired by IBM, died.  

 

NOTE THAT SUPERIOR COMPETITORS, HP’S NEW WAVE, NEXT’S NEXTSTEP, ARGUABLY TI’S PRODUCT, NINTEX, DOZENS OF OTHERS FAILED BECAUSE THEY WENT HORIZONTAL 

 

(Source:  Innovator’s Solution)

457

Narrow Focus on Extreme High Value for Customer

 

Whole Product Solution

 

Industry Expertise

 

Lost Their Way

Seibel – Pioneered CRM Applications.  MIXED - GOOD AND BAD  

 

  1. Saw half dozen apps for No Code Solutions 
  2. KEY WAS INTEGRATION ALMOST COMPLETELY PREPACKAGED 
  3. Ted worked for them five years, was #3 in the World in Sales 
  4. Success with implementation partners – why mixed 
  5. Selected partners based on industry strength, number of implementors / developers 
  6. KEY ITEM WAS PARTNERED WITH DELOITTE TO VERTICALIZE SOFTWARE.  THIS MADE DELOITTE HANDS DOWN THE PREFERRED IMPLEMENTATION PARTNER. 
  7. IMPLEMENTATIONS DRAMATICALLY BETTER WITH THESE SPECIALIZED FOCUSED IMPLEMENTATION PARTNERS (AS OPPOSED TO SOMEBODY WHO COULD JUST SELL THEIR WAY INTO WINNING THE IMPLEMENTATION BUSINESS) 
  8. QUICK HITS, SMALL WINS / KEEPING MOMENTUM AND FUNDING GOING WAS EVERYTHING 
  9. POSITIVES: 
    1. Great Tech 
    2. Verticalized the product 
    3. Already had interfaces working, ready to go 
    4. Picked good Big 4 partners with vertical strength 
  10. NEGATIVES: 
    1. Tom Siebol 
    2. Cocaine  
    3. Weekly call – micromanaging any deal over $1 million – worst issue – why he left 
    4. Was overfocused on Chase and Citibank – the big guys  
    5. Story about US Bank, customer who knew Tom Siebol’s problems and refused to meet with him 
    6. Numerous lawsuits 

 

(Source:  From former #3 sales exec in the world) 

 

456

Focus on The Pragmatic Buyer

 

Extreme High Value for Customer

Lessons from Microsoft

 

Microsoft’s position in history and size make it unlikely that its success will ever be replicated, but here are some lessons we can take away

 

Microsoft was built in large part on this concept:  “Customer funds initial software, we productize the software – make ready for large scale sales - sell big value for low cost” – e.g. MS Office.  Note There are significant costs to productize the software – making it ready for large scale sales.

Focus on pragmatic customers:  Low cost, high value software. 

 

SPECTACULAR WIN FROM THE SHIFT TO NON-PROPRIETARY HARDWARE.

 

(Source:  Former employees, TIA Research)

455

Narrow Focus on Extreme High Value for Customer

 

Authority / Responsibility Matching

 

Controlling Interference from Investors, Board Members

Sernic:  Canadian Supplier of ERP Software to NGOS (Non-Govt. Orgs.)

 

  • Grew from $4 Million to 12.5 Million in 7 Years
  • Profitable, But Would Have Done Better With Different Customer
  • STARTED WITH MARKET / CUSTOMER ANALYSIS
  • Sell what Customer will buy
  • Improve, Replace People
  • Enormous difficulty and distraction from board member continually demanding “What about Block Chain?”

 

(Source:  Randy K, former CEO)

 

454

Narrow Focus on Extreme High Value for Customer

 

ACI WORLDWIDE, INC., 2023 TIA Research Winner  100% ROI in six years

Complex FINANCIAL SERVICES software for transaction clearing, security between banks.  

 

- Large invested capital

- Very strong profits while growing revenue

- Unusually low market value

- CAUTION:  Possible problem period with customer hostility but has new CEO 

 

(Source:  TIA Research)

 

453

Focus

 

Barely A Winner, Lost Their Way

ALLIANCE DATA SYSTEMS CORP, 2023 TIA Research Winner   100% ROI In Seven Years

 

Credit card, loyalty and marketing services

 

- $25 Billion Valuation Down To $7 Billion.  Exited business, remainder now Bread Financial

- $3.5 Billion In Capital Tied Up For 10 Years+

- Startup unit for deregulated gas and electricity billing – ended losing in big litigation

(Source:  Tom Ingram Worked There Two Years)

 

452

Profitable from Day 1

 

Narrow Focus on Extreme High Value for Customer

 

 

 

CHECK POINT  SOFTWARE TECHNOLOGIES LTD  2023 TIA Research Winner  100% ROI in four  years - STRONGEST PROFITS IN STUDY

 

CYBER SECURITY SOFTWARE   

High invested capital

Huge valuation

 

(Source:  TIA Research)

 

451

 

Narrow Focus on Extreme High Value for Customer

 

Whole Product Solution

 

Services Sells Software

INTUIT INC  2023 TIA Research Winner  100% ROI in 4 Years, 200% ROI in 9 Years

 

FINANCIAL SERVICES / TAX SWR

 

NOTE:  Added services!

 

(Source:  TIA Research)

 

450

Whole Product Solution

 

Effective Reseller Industry Niches

Progress Software:  Big Win:    Spectacular Market Value On Low Invested Capital Due To Repayment Of Debt

No-Code / Low Code Software Development Tools

Strong industry focus programs for resellers

Source:  TIA Research 2009 Revised, interview with former executive, site visit

 

Whole Product Solution

 

Narrow Focus on Extreme High Value for Customer

 

Micros:  Big Win -  Spectacular Market Value On Very Low Invested Capital

30%+ growth some years with significant acquisitions

Strong profits, avg 12%

Point of sale systems and terminals for hospitality industry

Source:  TIA Research 2009 Revised

 

 

 

 

 

 

 

 

 

 

 

 


 

 

 

*Contact Us in Dallas, Texas, USA at tom@tomingraminc.com 

or 972-503-9287.

*Success stories, client quotes, estimated costs and benefits, names, faces, etc. are derived from actual projects

but may have been altered for simplicity, teaching purposes or to protect confidential information.

Names and faces are disguised to maintain privacy.  Contact us for details before making any purchase decision.