The Narrative Memo: Template and Method

Six months for a handstand, a week for a memo

In his 2017 letter to shareholders, Jeff Bezos tells the story of a friend who wanted to learn a perfect free-standing handstand. She hired a coach. In the first lesson he told her that most people think they can master a handstand in about two weeks. “The reality is that it takes about six months of daily practice. If you think you should be able to do it in two weeks, you’re just going to end up quitting.” ¹

Bezos uses it to explain why some Amazon memos are brilliant and others are not. The writers usually know a great memo when they see one. What they get wrong is the scope: they believe a six-page memo can be written in a day or two, “when really it might take a week or more!” ¹

A narrative memo is a short document in full sentences that sets out one decision, the real options, the evidence and a recommendation. Everyone in the meeting goes through it in silence before anyone talks, so the group argues about the substance instead of a presenter’s delivery.

Amazon is its best-known user: “We don’t do PowerPoint (or any other slide-oriented) presentations at Amazon. Instead, we write narratively structured six-page memos. We silently read one at the beginning of each meeting in a kind of ‘study hall.’” ¹

The thesis of this page: a memo beats a deck because full sentences won’t let you hide a missing “because.” A slide can put a conclusion next to three bullets and let the audience assume they connect. A paragraph has to say how they connect, or the gap shows. The week of writing is where the thinking happens. The silence at the start is what makes the room argue with the idea instead of the person.

A slide can hold a conclusion without its reason. A sentence can’t.

When a memo is worth the week

A week of writing is expensive, so spend it on decisions that are expensive to reverse. Bezos’s 2015 letter separates one-way-door decisions, consequential and irreversible or nearly so, which must be made methodically and carefully, from two-way-door decisions that high-judgment individuals or small groups can make quickly ². He warns that large organizations drift toward using the heavy process on most decisions, including reversible ones ². A memo for every decision is that drift in document form.

• Write a memo for entering or exiting a market, a large capital commitment, an acquisition, a reorganization, a pricing model change, or any decision several units must carry out together.
• Skip the memo for reversible operating choices, anything one owner can decide this week, and status updates.

The memo is one piece of a larger process: who decides, against what criteria, and how dissent gets handled. Our decision-making framework covers the rest.

The template

Bezos’s letter publishes no template. The one below is ours, built for strategy decisions: ten sections plus a header and an appendix, in this order. Six pages is the length Bezos describes. Most single strategy decisions fit in three or four, which the length guides assume.

Two sections carry most of the weight. The answer goes near the top so every later paragraph can be tested against it. The options section is where weak memos give themselves away: if the alternatives are straw men, the decider is being asked to approve, not to choose.

Writing it in a week

1. Agree the questionSpend 15 minutes with the decider to agree the question, the deadline and the criteria before any analysis.
2. Write the answer firstDraft sections 1, 2 and 10 on day one as a hypothesis. The analysis tests it.
3. Gather and date the factsFill section 4 and mark each fact as measured or estimated, with its source.
4. Build real optionsWrite at least three options, including doing nothing, and score them on the agreed criteria.
5. Get it attackedShare the draft with two colleagues and ask them to argue against the recommendation, then answer the best objection in section 8.
6. Rest, then cutPut it away for two days, then cut every sentence the argument does not need.

Step two borrows from hypothesis-driven problem solving: write the answer first so the work has something to test. The last two steps follow the rhythm Bezos describes. Great memos “are written and re-written, shared with colleagues who are asked to improve the work, set aside for a couple of days, and then edited again with a fresh mind” ¹. He also calls the writing teamwork: someone on the team needs the skill, but it does not have to be you ¹.

Five rules keep the draft honest:

• Full sentences in the argument. Bullets let you list points without saying how they connect. Tables are fine for comparisons.
• Numbers, not adjectives. Write “gross margin fell from 31% to 26% in six quarters,” not “margins are under pressure.” A weighted trade-off matrix keeps option scoring just as explicit.
• Concrete language. The writer knows what “operational excellence” means in this case; the reader does not. Chip and Dan Heath call this the curse of knowledge ³. Name the customer, the product, the plant.
• Every fact dated and sourced. A reader should be able to check any number without asking.
• State the best counterargument. Put the strongest case against your recommendation in its strongest form, and answer it.

Running the silent start

The memo changes the meeting. Hand it out at the start and give everyone silent time to go through it, the “study hall” Bezos describes ¹. Allow about twenty minutes for three or four pages. It will feel strange the first time. Twenty people sitting in silence looks like a meeting that has failed to start. It is the only part of most meetings where everyone is working on the same thing.

• Nobody presents. The team that wrote it answers questions only.
• Ask the decider to speak last, so the room hears the other views first.
• End by stating the decision in the words of section 10, or the specific information still needed and when it will arrive.
• Write the decision, the reasoning and the owner into the decision log the same day.

Why silence and not a pre-read? Pre-reads often get skimmed on the way into the room, so the discussion splits between people who went through the whole thing and people who saw page one. The silent start puts everyone on the same page, literally, before anyone speaks.

A filled-in example

Look at sections 7 and 8. The memo names the one condition the recommendation depends on (the August build-out) and the date that would trigger the fallback (June 15). A deck would have shown a green timeline. The sentence makes the bet visible.

The fair objection: nobody has time for six pages

The strongest objection comes from executives who sit through eight meetings a day: we cannot each spend a week writing and an hour going through memos. Close behind it is a sharper one. A polished memo can be advocacy in a nicer suit. Good writers win arguments they should lose, and the prose hides the same weak logic a deck would.

The time objection is mostly a misunderstanding of where the time goes. The reading happens inside the meeting, so it costs nothing extra. The writing week replaces the slide-building week most teams already spend; it just spends it on thinking instead of formatting. And the memo is for one-way doors only. A team that writes a memo for every decision has the wrong rule, not the wrong format.

The advocacy objection is real, and the template is built against it. Real options, criteria agreed before scoring and a section on what would change your mind make it hard to win on style. Amazon adds one more guard: by tradition, authors’ names never appear on the memos; the memo is from the whole team ¹. Bezos also concedes the standard is hard to specify; readers “know it when they see it” ¹. That is why the rest-then-cut step matters. It is the closest you can get to reading your own memo as a stranger.

If you are bringing memos into a slide culture, do not announce a new policy. Pick one decision meeting you own, send the memo with a note that the first twenty minutes will be silent, and let the quality of the discussion make the case. When the reader you need to convince is your own boss, our page on managing up covers how to sell the recommendation itself.

Common mistakes

• Burying the answer on page three. The reader spends two pages guessing where you are going instead of testing whether you got there.
• Options that are not real. The decider is asked to approve rather than choose, and learns nothing about the trade-off.
• Writing it in a day. Bezos’s point is that teaching the true scope, a week or more, by itself improves results ¹.
• Using it for a two-way door. The week is wasted, and the team learns that memos are bureaucracy.

The handstand coach’s advice was not about handstands. It was about telling people the truth about how hard something is, before they start. The memo works the same way. It tells the writer, and then the room, exactly how hard the decision really is.

Questions

What should a decision memo template include?

At minimum: the question being decided, the recommendation stated up front, why the decision is needed now, the facts with sources, three or more real options including doing nothing, the criteria and how each option scores, risks and what would change your mind, next steps with an owner, and the exact decision you are asking the meeting to make.

How long should a narrative memo be?

As long as the argument needs and no longer. Amazon’s practice, as Jeff Bezos described it in his 2017 shareholder letter, is a six-page memo. Most single strategy decisions fit in three or four pages, with supporting data moved to an appendix that nobody needs in order to follow the argument.

How does Amazon use six-page memos?

In his 2017 letter to shareholders, Jeff Bezos wrote that Amazon does not use slide presentations, writes narratively structured six-page memos, and reads one silently at the start of each meeting. He said great memos take a week or more, are rewritten and improved by colleagues, and carry no author names because the memo is from the whole team.

Who should write the narrative memo?

The person or team recommending the decision, not the decider. Bezos’s letter treats memo writing as teamwork: someone on the team needs strong writing skill, but it does not have to be the person who owns the recommendation. The decider agrees the question and criteria early, then tests the memo in the meeting.

Where to go next

• Getting a Room to Decide: A Decision-Making Framework
• Managing Up: Selling a Recommendation to a Skeptical Executive
• Strategy Narrative Builder (free outline)
• Decision Log
• Weighted Trade-Off Matrix (free grid)
• Reversibility / One-Way-Door Assessment (free scoring sheet)