A strategy that cannot answer hard questions is not a strategy. It is a mood.
When Your Strategy Is Just Someone Else's Framework in a Slide
A real product strategy makes specific bets that cost something. It names what you will not do, which customers you will not serve, and which metrics you will deliberately not optimize. If your strategy document could describe any competitor in your space, it is not a strategy.
A real product strategy makes specific bets that cost something, it says what you will not do, which customers you will not serve, and which metrics you will deliberately not optimize.
The Slide Everyone Has Seen
Picture the strategy review. The deck opens with a vision statement, bold, aspirational, uncontroversial. The next slide shows a two-by-two. Then comes the "strategic pillars" slide: three or four words in rounded rectangles, each pointing to a generic outcome. Delight users. Drive retention. Expand addressable market.
No one in the room disagrees. That is the problem.
A strategy that earns universal approval in a review meeting has, almost certainly, said nothing. It has described a direction so broad that it imposes no cost on anyone present. And a strategy that costs nothing guides nothing.
What Makes Something a Strategy-Shaped Object
A strategy-shaped object looks like a strategy. It has the right sections, the right vocabulary, the right slide count. It references the competitive landscape and names the target customer. It probably took a week to write and four rounds of revision to soften.
What it does not do is make a bet.
A bet has a cost. It means you have chosen to be worse at something in order to be distinctly better at something else. It means a specific group of people, a competitor, a customer segment, a stakeholder, will be disappointed by your choice. If no one is disappointed, no choice was made.
The tell is this: substitute your company name with your closest competitor's name in the strategy document. If the document still reads coherently, you have written a strategy-shaped object.
Three Tests for a Real Strategy
These are not steps in a process. They are diagnostic questions. A strategy that fails all three is a hiring pitch, not a product direction.
Test 1: Name three things the strategy rules out.
Not "we will deprioritize X", that is a roadmap deferral. The question is whether the strategy explicitly commits to never serving a use case, never optimizing a metric, or never entering a segment. The ruling-out must be costly enough that a stakeholder would push back if they read it.
If you cannot name three, the strategy has not made a choice.
Test 2: Identify the specific stakeholder it will disappoint.
Every real strategic choice has a constituency who loses. A decision to serve small teams over enterprises disappoints your enterprise sales lead. A decision not to optimize for daily active users disappoints a growth-oriented board member. A decision to sacrifice breadth for depth disappoints the user who wanted one more feature.
If you read the strategy and cannot name who will be upset, and why, specifically, the strategy has avoided making anyone upset. Which means it has avoided making a decision.
Test 3: Would a competitor change their behavior if they read it?
A real strategy is a signal. It tells competitors where you are building a moat, which segments you are concentrating on, and which battles you have decided not to fight. A competitor reading your real strategy should feel something: relief (you are not coming for them) or threat (you are coming directly for them).
If a competitor reads your document and has no new information about your intentions, the document contained no intentions.
Side-by-Side: Same Business, Two Documents
Both examples below describe the same hypothetical product, a B2B project management tool for engineering teams in India.
| Dimension | Strategy-Shaped Object | Real Strategy |
|---|---|---|
| Vision | Be the most loved project management tool for engineering teams | Be the tool engineering leads at 10-100 person Indian product companies choose over spreadsheets and Jira, not because we have more features, but because we never require a dedicated administrator to maintain |
| Who we serve | Engineering teams of all sizes who want better collaboration | Engineering leads at Series A and Series B product companies who own tooling decisions and have no dedicated project management operations person |
| What we will not do | (absent) | We will not build features that require more than one person to configure. We will not serve companies that need compliance reporting. We will not optimize for enterprise procurement cycles. |
| What we optimize for | Engagement, retention, and NPS | Time-to-first-useful-workflow for a solo engineering lead. Not DAU, not seat expansion, not feature breadth. |
| Who will be disappointed | (absent) | Our enterprise sales team. Every customer who wants a Gantt chart. Every investor who wants to see us compete with Jira on features. |
| Competitor signal | Competes broadly in the project management space | Ignores the top of the market and builds where Jira is architecturally incapable of being simple |
The strategy-shaped object on the left could describe Linear, Jira, Asana, Notion, or any project management tool that has ever existed. The real strategy on the right names a specific company size, a specific persona constraint, a specific metric to not optimize for, and the specific stakeholders who will object.
One of these documents is usable when a hard decision arrives. The other is not.
The Netflix Example
In 2007, Netflix entered streaming with a catalog that was, by any honest measure, inferior to cable. It had fewer titles, no live sports, no news, no day-of-air releases. Those were not gaps waiting to be filled, they were deliberate exclusions baked into the strategy.
Reed Hastings and his team made a specific bet: the future of video consumption would be on-demand and catalog-deep, not live and broad. They chose not to compete on breadth with cable. They chose not to serve the user who wanted to watch the game tonight. Those choices disappointed specific people inside the company and specific investors who wanted a bigger total addressable market story.
The result was a product oriented around something rather than everything. Every engineering decision, every content acquisition, every user experience choice had a filter: does this serve the person who wants to watch what they want, when they want, without a schedule? If not, it was out of scope.
That filter, derived from costly exclusions, is what made Netflix's strategy usable. When a decision arrived, the strategy gave an answer. That is the only test that matters.
Compare that to a strategy document that says "we will deliver the best streaming experience for entertainment lovers." That document gives no answer to any hard question. Should you bid on live sports rights? The document cannot say. Should you localize aggressively for India before solidifying the United States? The document cannot say. Should you sacrifice streaming quality on low-bandwidth connections to serve rural markets? The document cannot say.
A strategy that cannot answer hard questions is not a strategy. It is a mood.
The Judgment Turn
Most strategy decks are written to get approval, not to make decisions.
This is not cynicism. It is a structural reality. Strategy reviews involve stakeholders with conflicting interests. A document that satisfies the growth team, the enterprise sales team, the design team, and the board is a document that made no choice between them. The approval process itself selects against real strategy.
The incentive is clear. A document that disappoints no one gets approved. A document that makes real bets gets questioned, pushed back on, revised until the costly parts are softened into something inoffensive. By the time it ships, it is a strategy-shaped object.
The problem surfaces six months later, when a hard decision arrives and the strategy provides no answer. Should we serve this enterprise customer who does not fit our persona? The strategy says "serve engineering teams." Should we build this compliance feature the big customer is asking for? The strategy says "drive engagement." No one knows what to do, so the default wins: say yes to the customer in front of you, say yes to the feature that closes the deal.
That is not strategy failure. That is the absence of strategy revealing itself.
How to Fix a Strategy Document Without Starting Over
If your current document fails the three tests, you do not need to rewrite it. You need to add the costly layer that was edited out.
For each strategic pillar or initiative in the document, force a constraint statement in this form: "To pursue this, we will not [specific action], even if [named stakeholder] asks us to."
The constraint statement has to name the action and name the stakeholder. "We will not build enterprise compliance features, even if our sales team brings us a six-figure annual contract that requires them" is a real constraint. "We will focus on our core customer" is not.
Run each constraint statement through Test 2: would this disappoint a specific person reading it? If yes, keep it. If everyone would nod along, it is not a constraint, it is decoration.
Then run the full document through Test 3: if a well-resourced competitor read this, would they have new information about where you are and are not building? If the answer is no, the document still has no signal. Add the signal, or accept that you are operating without a strategy.
Key Takeaways
- A strategy that earns unanimous approval in a review meeting has almost certainly said nothing. Consensus is a warning sign, not a quality signal.
- The diagnostic test is substitution: replace your company name with a competitor's. If the document still reads coherently, no real bet was made.
- Real strategy is defined by its exclusions. Name what you will not do, which customers you will not serve, and which stakeholder will be disappointed. If you cannot, the strategy has not made a choice.
- The Netflix streaming strategy worked because it had costly exclusions, no live content, no catalog breadth race, that oriented every product decision for years. The exclusions were the strategy.
- A strategy written to get approved will guide no one. It will be most useless at the exact moment it is most needed: when the hard decision arrives and there is no longer time to debate direction.