A list of use cases, and a budget, with a logo on it. That is not a strategy — it's your shopping list!
Michael Porter settled what a strategy actually is back in 1996. Not a list of things you will do. A coherent set of choices that fit together, and, in his words: the essence of strategy is choosing what not to do.
Roger Martin and A.G. Lafley said the same thing in Playing to Win. Most companies are not playing to win; they are playing to participate, spreading a bit of budget across every opportunity so they never have to make a hard call. An AI roadmap that says yes to everything is the purest form of playing to participate ever invented.
A real AI strategy answers ten questions: six of them are actual strategy (choices where a reasonable company could decide the other way and pay for it), and the other four are the machinery that decides whether any of it survives your organization. Most companies answer two, and usually the wrong two.
The 6 Choices That Make Strategy
1. What You Own (The Scrooge McDuck Question)
What do you already own that a competitor cannot easily copy? Your proprietary data, your distribution, your domain expertise, the customer relationships you have built over years. David Teece drew the line that matters decades ago: some resources are ordinary (anyone can buy them at the market price), and some are hard to replicate. Durable advantage only ever comes from the second kind.
With AI, the model is available to everyone for the same price, so the tools are ordinary by definition. The only durable edge is the asset feeding them, and that is almost always your proprietary data. A strategy built on assets anyone can buy is a strategy anyone can copy.
2. The Vision
Not "become AI-driven" — which means nothing, because nobody is choosing to become AI-ignorant. Roger Martin's test: if the opposite of your choice is stupid, it is not a strategic choice. Your vision has to be specific enough that a reasonable competitor could choose differently.
3. The Why
Why do you want to innovate with AI at all? Plenty of profitable companies should barely touch it. If your honest answer is that everyone else is doing it, you do not have a strategy; you have anxiety with a budget.
4. The Definition of Innovation
What does innovation mean for your specific business? Selling AI-generated insurance quotes is disruptive for an insurer and trivial for a tech company. Copying someone else's definition is how you end up solving a problem you do not have.
5. The Mix
How much of your effort goes to small iterative gains, and how much to risky long bets? Iterative is safer, and your competitors will copy it in a single quarter. Disruptive is slower, and it is where the actual advantage hides. Getting the ratio wrong is a fatal strategic error.
6. What You Will NOT Do (The Boundary)
This is Porter's whole point. A strategy with no boundary is not a strategy, because it demands no trade-off, and without trade-off there is no choice. The company that says no to nine tempting AI projects to fund the one that matters is doing strategy. The company saying yes to all ten is doing theatre.
The Machinery: 4 Disciplines of Execution
The Slide Deck Test
Count how many of the six choices your current AI deck actually makes, and how many of the four disciplines it actually has. Most decks make one or two choices, skip the rest, and have none of the disciplines. That is a roadmap wearing a strategy costume.
A roadmap tells you what you will build. A strategy tells you what you own, why it matters, in what order you will move, at what cost, and what you are willing to give up to do it.
Porter published his definition thirty years ago. Martin reinforced it a decade back. Neither of them had AI in mind, and it does not matter. A strategy was always a set of choices that fit together, built on something you have that others do not, and delivered by systems most people find too boring to build. AI did not change any of that. It just raised the cost of not having one.
Ps: If your AI strategy fits on two slides and makes no one in the room uncomfortable, it is not finished. A real one always takes something away from someone.
- Porter, M. (1996), "What is Strategy?", Harvard Business Review.
- Lafley, A.G. & Martin, R. (2013), Playing to Win: How Strategy Really Works, Harvard Business Review Press.
- Teece, D.J. (2007), "Explicating Dynamic Capabilities," Strategic Management Journal.
Stop making shopping lists. Make real strategic choices.
brainTerms.ai enforces trade-offs, detects conflicting assumptions, and stress-tests your unit economics across every business domain before you spend a single dollar.