Amazon Banned PowerPoint. Google Has a Memo Rule. What Does Your Company Do?

On January 21, 2003, while the shuttle Columbia was still in orbit, engineers briefed NASA managers on the foam strike its wing had taken at launch. The analysis was real; the format was a PowerPoint slide so crammed with nested bullets and abbreviations that its most alarming finding was nearly invisible. The investigation board later reproduced that slide in its report and remarked that it was easy to understand how a senior manager might read it and not realize it addressed a life-threatening situation — then issued the sentence that made presentation software part of an accident record: the Board views the endemic use of PowerPoint briefing slides instead of technical papers as an illustration of the problematic methods of technical communication at NASA.

Seventeen months later — June 9, 2004, 6:02 PM — Jeff Bezos sent Amazon's senior team an email with the subject line No powerpoint presentations from now on at steam. His stated reason had nothing to do with NASA and everything to do with the same diagnosis: the narrative structure of a good memo, he wrote, forces better thought and better understanding of what's more important than what, and how things are related — and bullet points in Word, he added, would be just as bad. Years later he called the ban probably the smartest thing we ever did.

Now audit your own calendar. How many of this week's meetings will open with someone presenting slides the room is seeing for the first time — the presenter with hours of preparation, the deciders with seconds per bullet? This piece covers what Amazon actually does instead, the same pattern discovered independently at Google, Bridgewater and McKinsey, the honest numbers on what meetings cost, and a four-step version any team can run next quarter.

What Slides Do to Decisions

Start with the measured misery. In the survey behind Harvard Business Review's Stop the Meeting Madness (Perlow, Hadley & Eun, 2017), of 182 senior managers across industries, 65% said meetings keep them from completing their own work and 71% called meetings unproductive and inefficient. Those numbers describe the meeting as an institution; the slide deck explains part of the mechanism.

A deck creates information asymmetry by design: the presenter spent hours arranging the material; the room encounters it cold, in real time, while also trying to listen. And the format flatters weak reasoning — a slide reading improve satisfaction → raise NPS → launch loyalty program looks like a causal chain while containing three unexamined arguments. Edward Tufte's essay The Cognitive Style of PowerPoint (2003) made the canonical case — the format's compression flattens relative importance and hides the connective reasoning — and the Columbia board's finding gave the case its grimmest exhibit. Bezos's 2004 email is the same critique in one line: presentations give permission to gloss over ideas, flatten out any sense of relative importance, and ignore the interconnectedness of ideas.

The Amazon Method: Narrative Plus Study Hall

Amazon's replacement has two parts, and the second is as important as the first. The six-pager: a narrative memo, complete sentences and paragraphs, no bullet points, laying out context, the proposal, the expected results, and the risks. And study hall: every meeting built on a memo opens with silent reading — everyone, including the most senior person in the room, reads the same document before anyone speaks. Bezos described the ritual publicly at the Forum on Leadership in 2018; his 2017 shareholder letter added the quality bar: the great memos are written and re-written, shared with colleagues, set aside and edited again — they simply can't be done in a day or two. The decision philosophy the ritual serves — three good decisions a day, and the Type-1 / Type-2 split that decides which ones deserve a memo at all — is set out in his own words here.

The design solves three problems at once. Information symmetry: discussion starts after everyone has absorbed the same reasoning, so meetings begin where informed people disagree rather than at the background briefing. Authorial rigor: prose cannot hide a logical gap the way a bullet can — the author discovers the hole while writing, before the room does. And permanence: the memo is the decision record, not a souvenir of it — the reasoning survives, searchable, which is what makes it decision documentation rather than theater. The deeper anatomy of the six-pager — and the serious accessibility critique it has earned — gets its own companion piece: The Amazon 6-Pager Has a Problem.

Convergent Evolution: Four Companies, One Principle

But Won't Writing Slow Us Down?

The objection writes itself: a memo takes days; a deck takes an evening; we don't have the time. Bezos's own operating system is the cleanest answer, because the man who imposed the writing rule also imposed the speed rules — most decisions should be made with about 70% of the information you wish you had, and when consensus fails, disagree and commit. The memo discipline governs how reasoning is presented, not how long deciding takes; Amazon paired maximum rigor at the point of presentation with deliberate impatience everywhere else.

The triage from the satisficing playbook completes the answer: written-first treatment is for the consequential, contested, hard-to-reverse decisions — the one-way doors — not for every status update. And note what the writing time actually buys: the hours the author spends are hours the ten-person room does not spend being briefed, which is why the trade nets positive on any meeting whose attendees' time matters. A meeting that could have been absorbed asynchronously probably should have been.

Implementation for Any Size Company

Where Argumentree Fits

The written-first principle, taken seriously, ends in structure — which is exactly what Bezos's own words describe: what's more important than what, and how things are related. An argument tree is that structure made explicit. The proposal is the root; the arguments for and against are visible nodes with evidence attached; ratings show where the group actually stands; and the whole object is the permanent, searchable record the memo was always trying to be.

It also removes the memo's asymmetries: contributors add and attack arguments asynchronously before the meeting, the discussion starts from a shared structure rather than a shared reading burden, and nobody's reasoning is trapped in a deck that vanishes when the projector turns off.

The Rule Behind the Rules

Amazon's ban, Google's docs, Bridgewater's recordings and McKinsey's pyramids are one rule wearing four uniforms: thinking is externalized in a structured, examinable form before the room convenes. The Columbia board showed what the opposite costs at the extreme; the HBR numbers show what it costs every ordinary week.

You do not need Amazon's scale to adopt the rule — you need one meeting, one brief, and the discipline to read before you argue. The slides were never the problem, exactly. The problem was letting the presentation substitute for the reasoning. Fix that, and the meeting fixes itself.