Most Product Strategies Are Manifestation With a Roadmap

The test in one sentence

A product strategy you cannot prove wrong is not a strategy. It is a wish with a roadmap attached. The single question that separates the two is whether you can state, in advance, the specific evidence that would tell you the strategy is false. If nothing could ever disconfirm it, you are not strategizing. You are manifesting.

Most strategy decks fail this test, and they fail it in a way that feels productive. They are full of conviction, market sizing, and confident language about why this wins. What they almost never contain is the one line that matters: here is what we would have to see to know we were wrong, and here is when.

The tell that your strategy is a wish

The signature of a wish is that it survives every outcome. The product takes off, and the strategy was right. The product stalls, and it was execution, or bad timing, or the market was not ready. Notice that no result, good or bad, ever lands on the strategy itself. It floats above the evidence, permanently confirmed.

Karl Popper drew the line here for science decades ago. A claim that no observation could ever contradict is not a strong claim, it is an empty one. The same logic applies to strategy. If "we will win by being more customer-centric" is compatible with winning, losing, and everything in between, it predicts nothing, so it can guide nothing. It only feels like a strategy because it is written in the grammar of one.

This is exactly the move that defines manifestation. The belief is never disconfirmed by failure, because failure is always reassigned to the believer, who did not want it hard enough. A 2023 University of Queensland study found that the people who held this belief most strongly were no more successful, but were measurably more likely to make risky bets and end up bankrupt. The unfalsifiable belief does not protect you from being wrong. It only stops you from finding out in time.

What a falsifiable strategy looks like

The most useful correction comes from Roger Martin, who calls it the most important question in strategy: what would have to be true. Rather than argue about whether a strategy is right, you reverse-engineer it. You list everything that would have to be true for it to work, across customers, capabilities, costs, and competitors. Then you find the assumptions that are, in his word, heroic. The ones you would not personally bet on. Those heroic assumptions are not a weakness in the analysis. They are your test list.

Martin's other instruction is the one builders skip. He says to treat those assumptions as a canary in the coal mine and check, regularly, whether they are still true. The day the world stops cooperating with your core assumption is the day to retool, not the day you run out of money and finally look.

Put plainly: a real strategy hands you a short list of things that, if they turn out false, mean you are wrong. A wish hands you a story that is true no matter what happens.

Rewriting wishes as bets

The fix is mechanical. Take each strategic claim and rewrite it until it can fail.

Wish (survives any outcome)	Bet (can be proven wrong)
We win by being more customer-centric	Customers will switch to us for our onboarding speed. If trial-to-paid stays under 8% through Q3, the wedge is wrong
We are building a platform	Third-party developers will ship 30 integrations within six months of the API. Under 10 and we are a product, not a platform
We are early, the market will catch up	If weekly active teams do not pass 200 by December, "early" was a polite word for "wrong," and we cut spend
AI-native is our moat	If retention is not 15 points above the non-AI incumbent by v3, the model is a feature, not a moat

The right column is uncomfortable to write, and that discomfort is the point. The moment you attach a number and a date, you can lose. You can also learn, course-correct, and survive, which the left column never allows.

Make the roadmap a list of bets

A roadmap is where strategy quietly turns back into manifestation. It starts as a set of bets and decays into a list of features that nothing can remove. Every quarter the same items slide forward, untouched by evidence, because no one ever wrote down the condition under which they get cut.

A roadmap item with no kill condition is not a plan. It is a feature you have decided to love. Give each major item the same treatment as the strategy above. What does this bet assume, what would tell us the assumption is dead, and what do we do then. A roadmap where nothing can ever be cut by data is a vision board with deadlines, and it will hold its shape right up until the runway ends.

The five-question filter

Run every strategic claim through this before you commit a single sprint to it.

1. What would have to be true for this to work?
2. Which of those assumptions is heroic, the one I would not bet my own money on?
3. What specific, observable signal would tell me that assumption is false?
4. By when should I expect to see it?
5. What exactly do I do if I see it?

If you cannot answer all five, you do not have a strategy yet. You have a feeling with a slide deck. Answer them, and you have something far more valuable than confidence: a strategy that is capable of telling you it is wrong while you can still do something about it.

The takeaway

Falsifiability is not pessimism, and it is not a lack of belief. It is the only mechanism that lets you discover you are wrong early enough to matter. The founders who survive are not the ones with the most unshakeable strategy. They are the ones who built strategies that could be shaken, and then watched, honestly, to see if they were. A strategy you can kill is the only kind you can trust.