Product Strategy Memo Template (Pyramid Principle)

Six pages maximum, written to be read in silence. The structure is Barbara Minto's Pyramid Principle: answer first, then the grouped arguments that hold it up, then the evidence under each. A deck lets a weak argument hide behind a transition. A memo has nowhere to hide, which is the point.

Length budgets are marked per section. When you run over, the cut order is: appendix candidates out of the body first, then situation detail, then evidence you are keeping for sentiment rather than load. Never cut the anti-bets or the "what would change my mind" lines, they are the credibility.

The memo skeleton

# [Product/company name] Strategy: [The answer compressed to a title.
"Win the mid-market by owning the [workflow], not by matching enterprise
features"]

**Author:** [Name]  **Date:** [Date]  **Decision needed by:** [Date]
**Readers:** [Who this is for. Write for the most skeptical name on the list.]

**The answer in one sentence:** [The entire memo, compressed. The reader
who stops here should still know what you are proposing and roughly why.
If this sentence takes you a day to write, that is the day the memo
needed. Minto's rule: the governing thought comes first, everything below
exists to support it.]

---

## 1. Situation (budget: half a page)

[What the skeptical reader already agrees is true. Market position, the
core numbers, what the product does well today. No argument here, only
shared ground. The test: every sentence should make the reader nod, none
should make them reach for a pen. If a sentence is arguable, it belongs
in section 4 with evidence under it, not here. First section to cut when
over budget, your readers lived this.]

## 2. Complication (budget: half a page)

[What changed or what is now wrong, stated plainly. The two or three
forces that make the current course unavailable: a competitor move, a
cost curve, a trend in your own data, a customer behavior shift. Each
complication gets one sentence of statement and one of evidence. Three
maximum. A list of seven complications means you have not decided which
ones matter, and the reader will decide for you.]

## 3. Question (budget: 3 sentences)

[The question the complication forces. Frame it as a real choice between
live options, not a corridor to your answer. "Given [complication], do we
[A] or [B], and what do we stop funding in either case?" A question with
only one possible answer tells the reader the memo was written backwards.]

## 4. Answer (budget: half a page here, arguments below)

[Your answer, stated fully, with the shape of the reasoning. "We do [X],
because [argument 1], [argument 2], and [argument 3]." The three argument
names here become the three section headers below. Minto's grouping rule:
the three arguments must be mutually exclusive (no overlap) and
collectively sufficient (together, they justify the answer). If argument
2 is really a sub-point of argument 1, merge them and find the real
second argument.]

## 4a. [Argument one as a full sentence, not a topic.
"The mid-market workflow is underserved and our data proves we already
win it when we show up"] (budget: 1 page)

[Open with the argument restated, then the evidence in three slots:]

**Data:** [The one or two numbers that carry this argument, with source
and date. One strong number beats five weak ones. If the number needs a
paragraph of caveats, it is not strong enough to lead.]

**Customer evidence:** [Two or three verbatim quotes, dated and attributed
by segment. Verbatim, not paraphrased. Paraphrase is where wishful
thinking enters memos.]

**Competitive:** [What competitors are doing or structurally cannot do,
with a source. "They cannot follow because [structural reason]" is strong.
"They have not done it yet" is not an argument, it is a countdown.]

## 4b. [Argument two as a full sentence] (budget: 1 page)

[Same three evidence slots: data, customer, competitive.]

## 4c. [Argument three as a full sentence] (budget: 1 page)

[Same three evidence slots. If you only have two real arguments, write
two. A padded third argument weakens the two real ones, because the
skeptical reader attacks the weakest point and generalizes.]

## 5. Anti-bets: what we are explicitly not doing (budget: a third of a page)

[Three to five items. Each: what we are not doing, in one sentence, and
why, in one sentence. This section does more alignment work per word than
anything else in the memo. It converts a hundred future hallway debates
into one paragraph, and it proves the strategy is a choice rather than a
wishlist. A strategy with no anti-bets is a budget request.]

## 6. Risks (budget: a third of a page)

[The premortem summary. Run the exercise before writing this section:
assume it is 12 months out and the strategy failed, have the team write
down why, cluster the answers. Report the top two or three failure modes
here, each with: the risk in one sentence, the early signal we would see,
and what we do about it. Then one line per major risk: "What would change
my mind: [the evidence that would make us stop]." This line is the
difference between a strategy and a position.]

## 7. The ask (budget: a third of a page)

[Exactly what you need decided, by whom, by when. Decisions, not
blessings. "Approve [headcount/budget/kill] by [date] so [consequence]."
If the ask is only "alignment," the memo is an FYI and should say so.
Most strategy memos die here, with five pages of argument and no
checkable request at the end.]

## Appendix (no budget, but rules)

[Methodology, full data tables, the complete premortem output, detailed
competitive teardown. Two rules. One: nothing load-bearing lives here, the
memo must stand if the appendix is deleted. Two: everything here is
referenced from the body, an unreferenced appendix page is a page you
were afraid to cut.]

QA before it ships

[ ] The one-sentence answer survives alone, with zero context.
[ ] Sections 1-3 contain no arguable claims. All argument lives in 4.
[ ] The three arguments do not overlap, and together they are sufficient.
    (Minto: mutually exclusive, collectively exhaustive.)
[ ] Every quote is verbatim and dated.
[ ] The anti-bets section exists and names real things people want.
[ ] Every major risk has a "what would change my mind."
[ ] The ask is a decision with a date, not a request for alignment.
[ ] Six pages or fewer. If not, cut in order: appendix candidates,
    situation detail, sentimental evidence.

The retelling test (final QA, do not skip)

Give the memo to someone smart who was not involved. Let them read it once. Take it away. Ask them to retell the argument from memory.

If the retelling survives, the structure is sound: answer, three reasons, the things you are not doing, the ask. If the retelling collapses into "it was about a lot of things," the pyramid is broken. Fix the structure, not the prose. Strategy travels through an org by being retold in rooms you are not in, and the retelling test is the only rehearsal of that journey you get to watch.