What Product Strategy Is: A Six-Question Working Method
Product strategy is the set of choices about whom a product serves, which problem it addresses and how the team expects to succeed. My working format puts six questions on one page: audience, problem, advantage, exclusions, the central assumption and a checkpoint. It is a decision aid, not a universal definition or a substitute for research. A useful test is whether the page helps the team resolve an actual trade-off about what to build next.
Why I emphasize choices
Calling strategy a high-level plan does not tell a team which trade-off to make. I prefer to make the choices explicit: what we will pursue, what we will defer and which evidence would change our mind. That is the purpose of the format below.
A longer strategy document can be useful too. The one-page constraint is a way to expose vague answers before they become a long deck.
The one page
Start with six questions. I aim for a short answer to each, with evidence and unresolved questions linked separately.
| # | The question | What to make specific |
|---|---|---|
| 1 | Who is this for? | A defined audience you can reach and learn from |
| 2 | What problem, in their words? | The sentence they would say to a colleague |
| 3 | Why might we win? | The proposed advantage and evidence for it |
| 4 | What are we not doing? | Explicit exclusions for the relevant planning period |
| 5 | What are we betting on? | An assumption whose failure would change the decision |
| 6 | How do we know by when? | Observable evidence and an agreed review point |
I pay particular attention to exclusions and assumptions. Naming what we defer makes trade-offs easier to discuss, while recording an uncertain assumption gives us something to investigate. These questions help expose the reasoning; completing the table does not by itself establish a sound strategy.
Filled in: a real 0-to-1 product
ILTY is an AI companion whose historical project account reports an eight-week concept-to-release period, five AI personalities, a native app, seven languages and a 226-page marketing site. Those are our self-reported project figures. The following applies the six-question format retrospectively; it is not a contemporaneous strategy document or a report of validated demand.
Who: people seeking a private space for reflective conversation. Problem hypothesis: wanting to express a thought without involving a friend. Proposed advantage: distinct companion personalities, including a more challenging tone. Not doing: no clinical claims, no therapist marketplace, no Android at launch. Bet: some people will value a companion that questions them. Checkpoint: a defined paid-conversion measure after launch, with its target and observation window agreed before evaluation.
The platform and feature exclusions narrowed the work. The companion's tone connects a brand choice with a product hypothesis. The delivery timeline alone does not prove that those decisions caused the result or that the commercial hypothesis succeeded.
Project decision: guided deployment at Apstra
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 design decision was to make the deployment workflow explicit rather than offer an unconstrained canvas. That is the strategy example: a specific audience and problem, with a deliberate refusal. Juniper announced its intent to acquire Apstra in 2020; the announcement does not establish a transaction amount or a causal link between our design and the acquisition. This article does not include a reproducible deployment-time study, so we do not attach a percentage improvement to the example.
What product strategy is not
Different from a roadmap. A roadmap communicates direction and planned work; it may use dates, outcomes or time horizons. Strategy explains the choices behind that work. What a product roadmap is compares formats and their trade-offs.
Different from a vision. A vision describes the desired future. Strategy selects a route toward it and should identify assumptions that can be tested. The useful review horizon depends on the product and decision.
Different from OKRs. Objectives and key results express goals and measures. Strategy explains why those goals deserve attention and which competing opportunities the team will defer.
Different from a backlog. A product backlog organizes possible work and can already reflect prioritization. Strategy supplies reasons for including, ordering or declining that work; a backlog can help express those choices without replacing their rationale. The audience and positioning decisions also inform brand identity work.
What are the four major elements of a product strategy?
One useful four-part summary is audience, customer value, differentiation and business goals. In my format I also make exclusions and a checkpoint explicit. There is no requirement that a strategy use exactly four or six headings; what matters is whether the choices are supported and actionable.
What are the four types of product strategies?
There is no single required list of four. A team might compete through cost, a distinctive capability, convenience or a narrowly served audience. These are options to investigate, not mutually exclusive categories or a promise of success. For a small team, I usually explore a narrower audience first because it can make research and scope easier to manage.
How do I create a product strategy?
For a first workshop, I suggest reserving two hours to draft the six answers. Adjust that working session to the team and evidence available. Treat unsupported answers as research questions; completing the page does not complete the research.
- Draft the audience and problem. Separate customer evidence from assumptions. If the language is only your own, plan appropriate customer research before treating it as established.
- Discuss advantage and exclusions with the team. These are useful places to expose disagreement, though disagreement can arise in any answer. Record unresolved choices and their owners.
- Review the assumptions across all six answers. Identify which uncertainty could most change the decision and how you could investigate it.
- Agree a checkpoint. Name the evidence, review date or triggering event, and the person responsible for evaluating it.
- Try the page against a live decision. If it does not help, inspect what is missing: evidence, specificity, authority or a genuine trade-off. A week without a resolved argument does not invalidate the strategy.
Choose a review cadence that fits the decision cycle. Revisit the page when new evidence challenges an assumption, priorities change or the team cannot use it to decide between competing work.
Who writes it, and what that costs
The person accountable for product direction should own the decisions, with input from customers and the people who build, support and sell the product. At a small startup that may be the founder. Hiring and consulting costs depend on scope, seniority and location; this article does not establish a market rate. Our own Fractional CPO engagement starts at $15,000/month and combines direction with design and delivery within the agreed scope.
If the key uncertainty needs user evidence, choose the smallest credible test. That might be an interview, prototype, manual service or MVP. Our $9,000 sprint is one scoped implementation offer, not the required next step for every strategy.
The one-paragraph answer
Product strategy makes choices about the audience, problem and approach to success. The one-page format here adds explicit exclusions, assumptions and a checkpoint so the team can use those choices in day-to-day decisions. Use the format to expose missing evidence, then adjust the detail and review horizon to the decision.
For the earlier trade-offs behind a feature list, see design as a strategic tool. The founder essay UI is cheap explains why interface output alone cannot settle those decisions.
Have a product idea?
Designed, built, and shipped in about 5 days.