Skip to content

For investors and M&A teams

AI due diligence: questions about a target’s operating model

Most AI due diligence looks at the technology: models, platforms, data assets. The book points to a different risk — whether a company’s operating model is clear enough for AI to scale on, and whether two operating models can become one. Its merger case study shows savings stalling until definitions and owners were fixed, and this page turns that lesson into questions to ask.

A note on framing: the book does not contain a due-diligence checklist. Using its seven questions as a diligence lens is our application of the book’s method, not a claim the book makes.

Run the AI-Readiness ScorecardGet the book on Amazon

Kindle edition · 99 pages · a 90-minute readRead a free sample on Amazon

From the book

What the book shows

  1. 01

    Savings stall when systems merge before meaning

    In the book’s (anonymized) case, a cross-border merger was built to capture an estimated $250 million in year-two savings. Eighteen months in, integration had stalled and the two management teams were accusing each other of using bad data. The book’s conclusion is stated as the authors’ view, not a cited statistic.

    “Most mergers fail to capture their promised savings because leadership attempts the sequence in reverse, unifying technical infrastructure before standardizing meaning.”
    Chapter 20Case Study: When Two Strong Companies Could Not Become One
  2. 02

    What unlocked it

    The integration lead halted the system consolidation roadmap. Leadership built one combined map of how both businesses operate, co-authored a single data dictionary with non-negotiable definitions, and made named executives accountable for each domain. These are one company’s results as reported in the book — not a benchmark.

    “The stalled $250 million savings target was fully unlocked within nine months of the structural reset, driven by a unified view of vendor terms and procurement.”
    Chapter 20Case Study: When Two Strong Companies Could Not Become One
  3. 03

    Ready on paper is not ready

    The book’s other case is a company with a $45 million data platform, more than 350 executive dashboards and automated forecasting — whose regional presidents presented different values for the same KPI, and which could not make a decisive move when a logistics disruption demanded a fast reallocation of capital.

    “On paper, it was the textbook definition of a company ready for the AI era.”
    Chapter 19Case Study: The Company That Looked Ready

Questions to ask

Questions to ask of a target — or of two companies about to become one:

  1. 1Can management show, on one page, how the business operates — independent of its systems and legal entities?
  2. 2Do the companies define customer, order and revenue the same way? Where they don’t, who decides?
  3. 3Is there a named business owner for each critical data domain, with real authority?
  4. 4Does the integration plan unify systems before it unifies definitions?
  5. 5Which AI initiatives in the plan trace to a business decision and an accountable owner?
  6. 6Run the seven questions separately with several executives: how far apart are their answers?

Where to start in the book

  • Chapter 20

    Case Study: When Two Strong Companies Could Not Become One — the merger that could not become one company until it did so structurally.

  • Chapter 19

    Case Study: The Company That Looked Ready — the company that looked ready from the outside.

  • Chapter 17

    Scaling AI on a Foundation of Clarity — why operating clarity is becoming a measure of enterprise value.

Get the book on Amazon

Kindle edition · 99 pages · a 90-minute readRead a free sample on Amazon

Book cover: Your data isn't broken. Your operating model is.

Also see

All roles: who the book is for →

About this page: our reading of Your data isn’t broken. Your operating model is. by Adam Ayres and Behrang Karimi, for investors and M&A teams. The book itself is written for the leadership teams of complex enterprises. Quotes are word for word from the book; its case studies describe anonymized companies, and their results are not benchmarks.