Writing as a Thinking Tool: Why Amazon's 6-Page Memos Beat PowerPoint Every Time

Jeff Bezos, in a 2004 internal memo, banned PowerPoint presentations at Amazon senior-leadership meetings. Instead, any proposal had to be submitted as a six-page, narratively structured memo. Meetings began with 20-30 minutes of silent reading of the memo, after which discussion started. The policy was controversial at the time and is now widely imitated across serious operating companies. The reason: PowerPoint-driven discussions produced systematically worse decisions than memo-driven ones, and Bezos had watched enough examples to know the difference was material.

The shift from slides to memos is often described as a quirk of Amazon culture. It isn't. It's a specific insight about the relationship between writing and thinking — an insight that applies to far more than executive meetings, and that has implications for anyone whose work involves making decisions under complexity.

What Bezos Actually Said

In the 2004 memo — itself written in his memo style, with striking economy — Bezos identified two specific pathologies of PowerPoint that the memo format corrects.

First: bullet points hide bad reasoning. A bullet point can be vague, incomplete, or logically disconnected from the bullets around it, and the visual format disguises the gaps. You read "Key customer insights" and then three bullets, and your brain completes the implied argument even though the bullets don't actually make the case. In a prose memo, you have to write the connective tissue — the argument that links one point to another — and the gaps become visible.

Second: the presenter's skill dominates the analysis's quality. A charismatic presenter can sell a weak idea. A less-comfortable presenter can undersell a strong one. The room's judgement is influenced by performance rather than substance. The memo format neutralises this — everyone reads the same words, and the question becomes whether the thinking is good, not whether the delivery was good.

Bezos's conclusion, in the memo: "The reason writing a good 4-page memo is harder than 'writing' a 20-page PowerPoint is because the narrative structure of a good memo forces better thought and better understanding of what's more important than what." The key phrase is forces better thought. The memo isn't a better output format. It's a better thinking tool.

The Mechanism: Writing as Thinking

The connection between writing and thinking is older than Amazon. William Zinsser, in Writing to Learn (1988), argued that writing was not primarily an output medium but a cognitive process — that the act of putting thoughts into sentences forced a clarity that mental thought alone did not require. Stephen King says something similar about fiction. Paul Graham has made the same argument for essays.

The mechanism: when you try to think through an idea without writing, your brain accepts vague, incomplete, or contradictory fragments because they don't need to be reconciled with each other in any formal way. When you try to write the same idea, the sentences have to connect. You have to decide which thought comes first. You have to articulate why one claim supports another. The act of writing forces a precision that pure thought permits you to skip.

This is why some ideas that feel clear in your head turn out, when you try to write them, to be confused. The writing reveals the confusion. You discover, three sentences in, that you don't actually know how point A connects to point B. The problem was there the whole time. The writing is what surfaced it.

For this reason, professionals who write regularly — not for publication, but as a thinking practice — develop a specific kind of clarity that professionals who don't write lack. The writing is the discipline. The essays, memos, journals are the byproduct.

How the Amazon Memo Is Structured

The specific memo format Bezos required isn't documented publicly in full, but enough has leaked or been confirmed to reconstruct it. A typical Amazon six-pager has:

• Executive summary — one paragraph, the headline claim and recommendation.
• Context and background — what's happening, why it matters, what's been tried.
• Analysis — the core reasoning, often with data tables, customer quotes, or specific scenarios.
• The proposal — what specifically is being recommended, stated clearly.
• Risks and trade-offs — what could go wrong, what alternatives were considered.
• FAQ — anticipated questions and answers, which serves as a pre-emptive dialogue.

The FAQ section is the most distinctive. It forces the author to anticipate objections and address them in writing rather than waiting to be asked. The discipline of the FAQ is where a lot of the real thinking happens — the author discovers, while drafting the anticipated questions, that their case has specific weaknesses they hadn't faced.

A six-page memo is longer than most executives want to write and much longer than most want to read. That's the point. The friction is a feature. If the six pages can't be written coherently, the idea isn't ready. If the six pages can't be read, the reader isn't really engaging with the proposal.

Why This Is Hard to Adopt

Most companies that try to adopt the Amazon memo format fail within a year. The specific failure modes are predictable.

1. The memos become memos-in-name-only

Someone writes "memo" at the top of what is essentially a PowerPoint deck translated into prose. The bullet points become paragraphs in name only — each paragraph one thought long, no connective reasoning, no forcing function. The benefit is lost.

The fix is actual discipline. A memo has to argue its case, not just list its conclusions. Reviewing a proposal against the question "is this actually reasoning its way to a conclusion, or is it listing points?" catches most of these degraded versions.

2. The pre-read gets skipped

Meetings adopt the memo format but don't actually spend the first 20 minutes reading it. People arrive having skimmed, or not read at all, and then discussion begins with everyone at different levels of preparation. The memo format only works if every participant has actually read the memo before discussing it. Without that, you're worse off than PowerPoint — a memo being verbally summarised is even slower than a deck.

The fix: genuine silence, for real reading, at the start of every memo-based meeting. The 20 minutes feel awkward. They're also non-negotiable.

3. The FAQ becomes performative

The most valuable section — the FAQ — becomes a place where the author lists softball questions and gives easy answers, rather than honestly anticipating the toughest objections. This degrades quickly in organisations where memos are seen as political documents rather than thinking tools.

The fix is cultural, and hard. In healthy memo cultures, the author's credibility is built by anticipating hard questions and answering them honestly, including "I don't know" where that's the truth. In unhealthy ones, the FAQ is theatre. The difference is visible quickly, and the leadership of the organisation has to police it.

The Personal Application

You don't have to work at Amazon to benefit from writing-as-thinking. The personal practice that makes the biggest difference for most knowledge workers is simple: for any important decision or complex idea, write the full reasoning out in prose before acting on it.

Not bullet points. Not an outline. Prose, in sentences, with paragraphs. The length doesn't matter — 500 words, 2000 words, whatever the argument needs. What matters is the discipline of forcing the reasoning to be coherent, sentence by sentence.

The specific pattern that works for me: before any significant decision, I write a two-to-four-page memo to myself. What am I deciding? What do I think the right answer is? Why? What are the strongest objections? What would change my mind? I write it in one sitting, usually 30-45 minutes. I don't share it with anyone. The output isn't for an audience. It's to make my own thinking legitimate to myself.

About 30% of the time, the memo reveals that my initial inclination was wrong. The writing surfaces a weakness in the reasoning that pure thought had smoothed over. The decision I make after the memo is different from the decision I would have made without it. This happens often enough that the practice has paid for itself many times over.

The Claim That's Easy to Miss

The deep claim about writing-as-thinking is not that written output is clearer. It's that writing is a kind of thinking you cannot do any other way — that there are specific cognitive operations that only happen when you're composing sentences, and that these operations are disproportionately valuable for the kinds of problems professionals face.

If this is true, the people who write regularly — for any purpose, to any audience — are doing cognitive work that people who don't write regularly are not doing. Over years, this compounds into a difference in judgement quality that's visible and substantial.

Bezos's insight wasn't about meetings. It was about the relationship between form and thought. The memo format works because it enforces a cognitive operation that other formats don't. Most of the value for Amazon was the thinking it produced, not the output it created. The same insight applies to personal practice: the writing you do, whether or not anyone reads it, is doing cognitive work that the non-writing alternative doesn't do. Worth taking seriously, even if you never publish a word.