What Product Strategy Actually Is: A One-Page Definition
Product strategy is the set of choices that decides what you build next and, more importantly, what you refuse to build. It fits on one page and answers six questions: who this is for, what problem you are solving, why you win, what you will not do, what you are betting on, and how you will know within a quarter whether the bet was right. Every definition on the first page of Google today calls it "a high-level plan," which is where most strategies go wrong: a plan lists what you will do, and a strategy commits to what you will give up. If your document has no sacrifice in it, you have a roadmap with a nicer cover. The test is simple. Show it to an engineer and ask what to drop from this sprint. If the page cannot answer, it is not a strategy.
Nine of nine definitions say the same four words
We pulled the live results page for "what is product strategy" on 4 September 2026. Google's AI Overview leads with "a high-level plan that outlines what a product hopes to achieve." Below it, Atlassian, ProductPlan, Aha, Productboard, Wikipedia, Amplitude, Product School and Built In all open with a variant of the same phrase: a high-level plan, a coordinated plan, a blueprint. Roman Pichler is the outlier, framing it as "a consistent set of specific choices," which is the useful version and the one almost nobody copies.
The consensus is not wrong, it is just unusable. A high-level plan describes a document's altitude, not its content. It gives you no way to tell a good strategy from a bad one, which is why so many teams have a strategy deck and still argue about what to build on Monday.
The one page
Six questions. If an answer needs more than two sentences, the thinking is not finished.
| # | The question | What a real answer looks like |
|---|---|---|
| 1 | Who is this for? | A named segment you could buy a list of, not "SMBs" |
| 2 | What problem, in their words? | The sentence they would say to a colleague |
| 3 | Why do we win? | A structural advantage, not "better UX" |
| 4 | What are we not doing? | Three named things you are refusing this quarter |
| 5 | What are we betting on? | The assumption that, if false, sinks the plan |
| 6 | How do we know by when? | One number and one date |
Line 4 is the one people skip, and it is the line that makes the page a strategy. Line 5 is the one that makes it honest: writing down the assumption that could sink you is uncomfortable, which is exactly why it belongs on the page.
Filled in: a real 0-to-1 product
ILTY is an AI mental-health companion we took from concept to shipped iOS product in eight weeks: five AI personalities, a native app, seven languages, and a 226-page marketing site. The strategy page behind it was six lines long.
Who: people who journal at 2am and will not book a therapist. Problem: "I need to say this out loud to something that will not judge me or tell my friends." Why we win: the companion confronts rather than soothes, which is the opposite of every gentle wellness app. Not doing: no clinical claims, no therapist marketplace, no Android at launch. Bet: people will pay for an AI that pushes back, not one that agrees. Checkpoint: a paid conversion rate we could read within one quarter of launch.
Notice what the "not doing" line bought. Cutting Android and the marketplace is what made an eight-week build possible, and the confrontational tone, which sounds like a brand choice, is the whole strategic bet. Both decisions came from the page, before a screen was designed.
Filled in: a strategy that survived an acquisition
We designed the Day 0 workflow for Apstra, an intent-based networking platform, across six years. The strategic choice was narrow and unglamorous: stop treating network design as a document to be authored and start treating it as a build order to be guided. That meant refusing the flexible, spreadsheet-like canvas enterprise buyers kept asking for.
The refusal is the strategy. Deployment time dropped 90%, onboarding became self-serve, and Juniper Networks acquired the company for $450 million. The full proof points are on the Fractional CPO page. A flexible canvas would have been a better demo and a worse product.
What product strategy is not
Not a roadmap. A roadmap is a sequence with dates. It answers when. Strategy answers why that sequence and not another. Teams that have only a roadmap re-litigate priorities every planning cycle, because there is no page to point at when someone asks why a feature is not on it.
Not a vision. A vision is where you are going in three years and is deliberately unfalsifiable. A strategy is falsifiable this quarter, which is what makes it useful and uncomfortable.
Not OKRs. Objectives set targets. Strategy explains why those targets and not others, and which work you will refuse in order to hit them. OKRs without strategy produce teams that hit every number and lose the market.
Not a backlog. A backlog is a list of everything anyone ever asked for. Strategy is the filter that decides which items are allowed to become work. The same filter decides what the brand has to say, which is why identity work starts with positioning rather than a logo.
What are the four major elements of a product strategy?
The common framing, and the one Google's AI Overview repeats, is four: target audience, customer value, differentiation, and business goals. It is a reasonable list and it is missing the two elements that make a strategy binding. Add the refusal (what you will not build) and the checkpoint (the number and date that tell you whether you were wrong). The four elements describe a market position; the six questions make it a decision you can be held to.
What are the four types of product strategies?
The textbook answer is cost leadership, differentiation, focus or niche, and quality. Those are competitive strategies borrowed from Porter and applied to products, and they are useful for classifying a business, not for deciding what to build on Monday. In practice startups pick from a shorter list: be cheaper, be faster, be narrower, or be the only one who does the hard thing. Narrower is the one that works most often with a small team, because it is the only one that does not require outspending someone.
How do I create a product strategy?
Book two hours, not two weeks. Write the six answers, and write them badly, then fix them.
- Draft line 1 and 2 alone. If you cannot describe the customer's problem in their words, stop and go talk to five of them.
- Fight over line 3 and 4 with your team. Why you win and what you refuse are the only two lines that generate real disagreement, which is why they generate real alignment.
- Write line 5 last. The bet is usually hiding in whatever you assumed while writing lines 1 through 4.
- Put a date on line 6. A checkpoint without a date is a wish.
- Test it against a real decision this week. If the page does not resolve one live argument in the first seven days, it is decoration.
Re-read it monthly, rewrite it quarterly, and rewrite it immediately when the bet on line 5 turns out to be false.
Who writes it, and what that costs
At a startup, the founder writes it. That is not a phase to grow out of, it is the job. It becomes a hiring question when the founder cannot hold direction, design, and delivery at once, and the honest numbers are these: a full-time Chief Product Officer costs $250,000 to $400,000 a year plus equity, and most fractional CPOs charge $8,000 to $20,000 a month for advice and a roadmap. Our own Fractional CPO engagement is from $15,000/month, and the difference is that the same person who sets the direction designs and ships it. A strategy nobody builds is a document; a strategy the author has to ship is a decision.
If the answer to line 6 is "we cannot know until something is in front of users," you do not need a strategy consultant, you need the smallest real product that can answer it. That is what an MVP is for, and we price the sprint that builds one at $9,000 flat.
The one-paragraph answer
Product strategy is the set of choices about who you serve, what problem you solve, why you win, what you refuse to build, what you are betting on, and how you will know within a quarter whether you were right. It fits on one page. Every definition on Google's first page today calls it a high-level plan, which is true and useless: the plan describes what you will do, and the strategy is the part where you decide what to give up. If your document contains no sacrifice and no date, it is a roadmap wearing a strategy's clothes.
Have a product idea?
Designed, built, and shipped in about 5 days.