Product Strategy Document Template

Start with everything that already claims to be strategy: the current draft, the board deck, last year's strategy and the company plan this one sits under. Keep the versions that disagree. Where two drafts contradict each other, you've usually found a decision nobody has made yet.

Then collect the evidence behind the numbers: market research, win and loss notes, customer interviews and product analytics. Next to every claim, write the date of its source, not the date of the slide that repeats it. A 2023 report quoted in a 2026 deck is a 2023 claim.

This template marks any source over 18 months old as stale. Old market sizing gets a second flag, since that number usually carries the most weight and gets rechecked least. If a claim cites a trade article, find the original research or drop it. When you need a fresh market estimate, build it from sourced figures instead of reusing an old headline.

Copy each strategic statement into one list, word for word, since paraphrasing quietly repairs weak lines. Roger Martin, co-author of Playing to Win, asks whether the opposite of your choice is stupid on its face¹. If it is, you haven't made a strategic choice. Nobody argues for a worse user experience, so "We will deliver an exceptional user experience" fails.

Martin's proof that an opposite isn't stupid is that rivals choose it and still succeed. A competitive positioning map shows who made which bet. Lines that fail are what he calls operating imperatives, things you must do just to keep up. Keep them on a separate list, and limit the real choices to three to five, the range he suggests.

Marty Cagan, founder of Silicon Valley Product Group, says he keeps finding companies with "a list of literally 50 major objectives or initiatives"² on a wall or spreadsheet. Their product teams, he writes, are committed to more than 100% of their time.

Each line that passes faces two more questions. Did someone you work with actually argue for the opposite? Write down their name and their case, because an alternative nobody championed is usually a straw man. And what would show the choice is wrong? Pick a number someone already watches, not "if the market shifts."

Next, read the surviving choices against each other and against the roadmap. In his 1996 HBR article What Is Strategy?³, Michael Porter describes how Continental Airlines launched Continental Lite to match Southwest's cheap point-to-point flights while staying a full-service airline. He calls this straddling. Continental lost hundreds of millions of dollars, and its CEO lost his job.

Product drafts straddle in familiar ways. A self-serve growth plan sits beside a buyer who runs a 90-day committee review, or "mobile-first" sits beside a product whose main users work at desks. Don't soften both lines until they fit. Decide which one the company is doing this year, judging by where money and people go, and move the other to the not-doing list.

In Porter's words, "The essence of strategy is choosing what not to do." Build the not-doing list from the alternatives that lost, and record who argued for each, its strongest case and the condition that would bring it back. Without that condition, a declined idea tends to get argued again at every planning meeting.

Every choice rests on beliefs about customers, competitors and costs that aren't settled yet. Martin calls "What would have to be true?"⁴ the most valuable question in strategy. In his experience, teams that disagree about the facts can almost always agree on the logic of a choice. Then they can agree in advance on the evidence that would settle it.

Write each assumption as a claim about the world, then give it a threshold and a date, so it can come back false. If the evidence is two years away, call it a review item, not a live test.

Recheck the list as evidence arrives. When something that had to be true stops holding, Martin says to revisit the strategy "right now," not at the next planning cycle. Run the cheapest tests first: a handful of discovery calls can settle the assumption behind a two-country expansion before any engineering starts.

The final document should be prose, not slides. Jeff Bezos wrote in Amazon's 2017 letter to shareholders that the company doesn't use slide presentations. It writes "narratively structured six-page memos"⁵, and a great one "might take a week or more." Prose forces the reasoning to connect: this buyer, therefore this product, therefore this way of selling. The template's two-to-four-page narrative follows this order:

1. 1

   Where you stand

   The situation in current numbers, with any stale figure dated or dropped.

2. 2

   The choices

   One paragraph each, naming the alternative it beat and why it lost.

3. 3

   The bets

   The two or three assumptions that would change everything, with their tests and dates.

4. 4

   What you won't do

   A short not-doing list, with what would bring each item back.

5. 5

   What's still open

   The questions this strategy doesn't settle yet, stated plainly.

Once the strategy holds, your product roadmap should follow from it, and RICE scoring can rank the work that serves its choices. To start from your own drafts, open this template in River, which tests each statement and dates each claim for you.