Digital download. All sales are final: no refunds or returns. Each format can be downloaded 3 times, and the links stay valid for 30 days. If you need more, reply to your confirmation email and we will sort it out.
About the AI Strategy Reference Model
Where should a company place its AI bets, buy or build, and how have the early movers fared? The AI Strategy Reference Model turns the ai strategy framework into a file you can work through, and then shows you the decision being made for real in 8 verified records with dated outcomes.
Two kinds of content, kept apart on purpose. The model is Stratrix's framework guidance, built from the free AI Strategy blueprint, organized decision by decision, and labeled as judgment. The worked records are verified fact: every one is cited to its sources and passed the same checks as every casebook record. Across the 8 records, which span 2013 to 2025, 8 carry dated outcomes; 6 of the decisions held, 1 were reversed, 1 held in part, and 0 are still too early to call.
Why it is worth your time
AI strategy is where the gap between the announcement and the outcome is widest, which makes verified records unusually valuable. The model covers the seven components from readiness to scaling and governance. The records are the early movers with outcomes on the books: Google shipping Bard before it was sure and then folding AI into search, IBM scaling Watson Health before the science held, Baidu's pivot to AI, Thomson Reuters and McDonald's deciding to buy the AI layer rather than build it, and Intel passing on a stake in OpenAI.
The components
An AI Strategy is the deliberate plan for how an organization will identify, develop, deploy, and govern artificial intelligence capabilities to create measurable business value. It is not a technology roadmap or an inventory of AI tools — it is a business strategy that defines where AI will create competitive advantage, how the organization will build the capabilities to deliver it, and what guardrails will ensure responsible use.
- AI Readiness AssessmentThe Organizational Reality Check
- Use Case PortfolioWhere AI Creates Value
- Data Foundation for AIThe Make-or-Break Layer
- AI Talent & CapabilitiesThe Human Architecture
- Responsible AI & EthicsThe Trust Architecture
- AI Operating ModelThe Delivery Engine
- AI Scaling & GovernanceFrom Pilots to Enterprise Capability
Each component in the file carries its description, what you have to decide, and the tables, lists and do's and don'ts that go with it. The file also carries 4 patterns, each with what it is best for, and 6 pitfalls, each with its symptom and its prevention.
Who reaches for it
The strategy lead running a ai strategy workstream who wants the structure and the precedents in one place. The consultant who has to turn a framework into a workshop and a deck by Monday. The founder or operator facing this decision who would rather see how it went for Alphabet (Google), IBM and Baidu than read another essay. And the faculty member who teaches the framework and wants cited cases to teach it with.
How people use it
In the workshop, the PowerPoint gives you one slide per component with what to decide, and one slide per record to show the decision made. In the memo, the Word file gives you the framework and the records with sources, ready to cite. In the model, the Excel lets you score your own situation against each component and sort the records by verdict. And in your AI workspace, the JSON gives the assistant the framework and verified precedents as context, so it reasons from records instead of inventing them.
What is in the download
A sample record
One of the worked records, free, exactly as it appears in the file. The other 7 verified records are in the purchased files.
| Company | Fork date | The fork | Options | Choice | Stated rationale | What happened | Judgment analyst | Sources |
|---|---|---|---|---|---|---|---|---|
| Alphabet (Google) Read the story → verified | 2023-02 | Ship a system that will sometimes be confidently wrong, or wait until it isn't |
| Ship under competitive pressure | A rival made waiting look like weakness, so Google crossed its self-imposed reputational-risk line and shipped | Latest · 2023-02-08 Alphabet market value: Alphabet shed about 7.7% of its value - roughly $100 billion | reversibility low verdict held | Google (official blog) (2023-02); CNBC (2023-02); CNN Business (2023-02); Google (official blog) (2024-02) |
Who is in the worked records
8 verified records across these companies, drawn from 6 decision types: Bet-the-Company, The Money Machine, The Fall, Adjacency Expansion, The Counterfactual and Ecosystem Lock-in.
- Alphabet (Google) ×2
- Baidu
- IBM
- Intel
- McDonald's
- ServiceNow
- Thomson Reuters
Questions
- What is in the AI Strategy Reference Model?
- The ai strategy framework as a structured model: 7 components, each with a description and what to decide, the option tables, the do's and don'ts, 4 patterns and 6 pitfalls. Then 8 verified decision records of Alphabet (Google), IBM, Baidu, Thomson Reuters, McDonald's, Intel and 1 more making this kind of decision, each with the fork, the options, the choice, the stated rationale, what it committed, what happened at one, three and five years, and every source. Five formats: Excel, JSON, PowerPoint, Word and PDF.
- How is the model different from the free blueprint on the site?
- The blueprint is the essay. The model is the working file: the same framework broken into components you can walk through in a workshop, a deck with one slide per component, a spreadsheet of what to decide, and JSON your AI workspace can reason from. The blueprint's company anecdotes are not in the file; the verified records replace them, and every one is cited.
- Are the records the same as the ones in the casebooks?
- Yes. Each record also appears in its own decision-type casebook and company casebook. The model collects the 8 that show this framework's decisions being made, in a curated order, so you do not have to find them across 6 casebooks.
- Is this a subscription?
- No. It is a one-time purchase, re-issued whenever a story a record rests on changes. The library license keeps every casebook and model current for a flat annual fee.