The Six-Page Memo vs the Deck: When Prose Beats PowerPoint

Write a narrative memo instead of a deck whenever the recommendation is complex enough that skipping a logical step would go unnoticed in bullet form. Amazon's six-page memo format works because full sentences force a writer to complete an argument, while a slide's fragments let missing evidence hide behind a confident bullet.

Quick Answer: Prose forces complete arguments; bullets let gaps in reasoning hide behind confident formatting. Write a six-page narrative memo for complex, high-stakes recommendations that need to survive scrutiny. Reach for a deck when you're live-persuading a room or walking through inherently visual data.

Why Amazon Banned the Bullet Point

Amazon eliminated PowerPoint from internal decision meetings because bullet points let a presenter's charisma paper over gaps that would be obvious in full sentences. A slide can jump from problem straight to conclusion without ever stating the reasoning in between; a memo has to write that reasoning down, in complete clauses, for someone else to check.

Jeff Bezos described the reasoning behind the switch in Amazon's 2017 shareholder letter: a well-argued, narratively structured memo forces better thinking and creates better shared understanding than a slide deck ever could, because a slide format simply doesn't require a complete thought. The logic here is structural, not stylistic. A bullet is a claim with its supporting argument removed — the writer knows why it's true, but the reader is asked to trust it rather than see it.

Edward Tufte made a version of the same case more forensically in his essay The Cognitive Style of PowerPoint, pointing to NASA's own internal slides as evidence. The Columbia Accident Investigation Board later found that a critical engineering risk — foam-strike damage to the shuttle's wing — had been present in a briefing slide, its urgency flattened by a crowded bullet hierarchy rather than argued out in prose. Nobody in that room had lied. The format itself had absorbed the warning.

The underlying argument, stripped down: a slide can present a conclusion with no visible reasoning between it and the evidence behind it. A memo can't. Every sentence has to connect logically to the one before it, in full view of whoever is reading.

Bullets conceal specific failure modes that prose exposes almost immediately:

• Missing causal links — "sales dropped" sits next to "increase marketing spend" with no argument connecting the two.
• Unstated assumptions — a projection with no visible base rate, growth assumption, or confidence range attached.
• Cherry-picked evidence — three supporting bullets with no visible counter-evidence, because there was no obligation to include it.
• Skipped alternatives — a recommendation presented as the only option, because slides rarely narrate the ones that got rejected.

This is why the discipline shows up before the meeting, not during it. A writer forced to render an argument in full sentences discovers its holes while drafting; a bullet lets the same hole pass silently onto the slide, where it's the room's job to notice — if anyone does. For a broader map of when PMs should write instead of talk, and vice versa, the complete guide to PM communication is a useful companion to this piece.

Most teams that try the format resist it at first, for a reasonable reason: a memo takes longer to write than a deck takes to assemble. That's the trade, not a flaw in it. Every narrative memo vs slides decision is really a decision about where you want the effort to go — into writing a defensible argument before the meeting, or into presenting confidently around whatever gaps the slide format let you skip.

Inside the Six-Page Memo: The Structure That Forces Rigor

A six-page memo works because its structure is fixed: context, goals, the data behind the recommendation, the recommendation itself, alternatives considered, and an appendix. Amazon veterans Colin Bryar and Bill Carr describe this shape in Working Backwards as deliberately unglamorous — the format's job is to surface gaps in logic, not to look polished.

Most teams adapt this into a repeatable template rather than treating six pages as a hard rule:

The fixed order matters more than the page count. Because alternatives considered gets its own section, a reader can't approve a recommendation without also seeing what was turned down — exactly the step a deck lets a presenter narrate around, live, without ever writing it down.

The summary paragraph is where the Pyramid Principle earns its keep: state the conclusion first, then let the rest of the memo supply the argument underneath it, instead of building toward a reveal on page five. Framing that opening paragraph with an SCQA structure — situation, complication, question, answer — keeps the first page from reading like throat-clearing before the actual ask.

The PR/FAQ Variant for New Products

Amazon's product teams often write a specific six-pager variant called the PR/FAQ: a mock press release for the not-yet-built product, followed by the internal and external questions it will raise. It forces the same discipline from a different angle — writing the customer-facing outcome before the internal plan — which Working Backwards describes as harder to fake than a roadmap slide, since a press release has to name one real customer benefit in a single sentence.

The Silent-Reading Ritual: Why Quiet Beats a Pitch

Amazon meetings built around the six-pager typically open with roughly 20 to 30 minutes of silent reading in the room, rather than assuming everyone pre-read. This ritual matters because it guarantees every participant has actually engaged with the argument before anyone speaks, closing the gap between who prepared and who merely showed up.

Pre-reading sounds efficient and rarely happens in practice. A document emailed the night before competes with everything else in an inbox, so most attendees arrive having skimmed the first paragraph at best. Silent reading in the room removes that excuse — for half an hour, the memo is the only thing anyone in the meeting is doing.

It also changes what the discussion afterward looks like. Comments and questions get written in the margin as they're read, not composed as a rebuttal to whatever the loudest person in the room said first. The room ends up debating the argument as written, not a paraphrase of it delivered from memory.

That discipline pairs naturally with writing the bottom line up front: a reader given the conclusion in the first paragraph spends the rest of the memo checking the argument against a claim they already know, rather than hunting for what the point even is. It's a small structural choice that removes an entire category of confused, meandering discussion.

When the Deck Is Still the Right Call

A deck outperforms a memo whenever the goal is live persuasion, a room that needs a shared image to react to together, or an audience that won't sit through silent reading first. Slides compress a chart, a demo, or one dramatic number into a moment a whole room can react to at once — a memo can't replicate that shared visual beat.

The two formats aren't rivals so much as tools calibrated for different rooms:

Notice that the deck's strength is exactly the memo's blind spot, and vice versa. A board meeting that needs to watch a revenue chart move in real time is a poor candidate for 30 minutes of silent reading, just as a memo is a poor container for image-heavy customer research that needs to be seen, not described.

Nancy Duarte, whose firm has built presentations for hundreds of executive teams, makes a version of the same argument from the opposite direction in Resonate and slide:ology: a deck earns its place when the goal is a shared visual or emotional moment, not a substitute for written analysis. Her guidance for that context mirrors Amazon's from the other side — one idea per slide, minimal text, a visual built to be glanced at rather than read line by line.

Reach for a deck, specifically, when:

• You're presenting live to an audience that expects spoken narration, like a board or an external partner unfamiliar with the memo format.
• The core evidence is inherently visual — a trend line, a map, a product screenshot — and prose would just describe a picture instead of showing it.
• Time is fixed and short, with no room for silent reading before the conversation even starts.
• You're building consensus in real time, where the ability to skip ahead, backtrack, or riff off a reaction matters more than a fixed, linear argument.

How to Write Your First Narrative Memo

Start by writing the one-paragraph summary and the recommendation before touching any supporting data or evidence. If you can't state the ask clearly in five sentences, the memo isn't ready to draft yet, and no amount of appendix detail or polish will fix that underlying problem.

1. Write the summary paragraph first. State the recommendation and the ask before writing any supporting section — if it takes more than five sentences, the recommendation likely isn't settled yet.
2. List the alternatives you rejected, on purpose. If you can't name at least two, you probably haven't looked hard enough for them.
3. Write the data section in prose, not a wall of raw numbers. Translate every figure into what it means for the decision, not just what it measured.
4. Have someone else silent-read it before the meeting. Their confusion in the margins is more useful feedback than their verbal agreement afterward.
5. Cut, then cut again. Six pages is a ceiling, not a target — a tighter memo that survives scrutiny beats a longer one that merely fills the page count.

The most common mistake first-time writers make is treating the memo as a report on work already decided, rather than an argument still being tested. A memo that only justifies a conclusion someone already reached reads as defensive, and a sharp reader in silent-reading mode will notice the difference between a case being made and a case being rationalized. Writing the alternatives section honestly — including the option you personally favored less — is usually the fastest way to avoid that trap.

The evidence section is usually where a memo's advantage over slides shows up most clearly, because that's where structured analysis actually lives. If the recommendation rests on a Jobs to Be Done analysis of why customers are hiring — or firing — a product, prose can walk through the competing forces pushing and pulling on that decision in a way a single insight bullet never could.

The same holds for a mapped customer journey emotion curve: a chart can show where satisfaction dips, but only prose can argue why it happens and what specifically closes the gap.

Writing the Memo Somewhere Built for Prose

Most tooling built for PMs defaults to fragments — a roadmap board, a slide template, a ticket description field. Prodinja's Library is built around the opposite assumption: that a recommendation of this size deserves a full narrative document, not a summary collapsed into bullets. Drafting the six-pager there means the complete argument — context, data, alternatives, appendix — lives as one document you can revise as the case sharpens, instead of getting flattened into slide-sized fragments somewhere along the way.

Key Takeaways

• Bullets delete the argument, not just the words. A slide can present a conclusion with no visible reasoning behind it; a memo forces the writer to supply that reasoning in full sentences.
• The six-pager's fixed structure — context, data, recommendation, alternatives, appendix — is the point. It guarantees a reader sees what was rejected, not just what was chosen.
• Silent reading closes the prep gap. Twenty to thirty minutes of reading in the room means everyone engages with the actual argument before anyone speaks.
• Decks still win for live, visual persuasion — board pitches, product demos, and anything better shown as a single chart than argued in prose.
• Write the summary and alternatives sections first. If the recommendation can't survive being stated in five sentences, it isn't ready for six pages either.
• Structured evidence like a JTBD analysis or a journey map deserves prose, not a single bullet, whenever the reasoning behind the finding is part of the case being made.