Writing is culture
Why the best organisations write to decide, not just to document
Welcome back for the June subscribers edition. Thank you for your continued support.
In last month’s article, I walked through a process for extracting your implicit leadership principles from your own writing. The key insight, borrowed from Larry McEnerney’s lecture on writing, was that you don’t think first and then write: the process of writing is the thinking.
Everything you’d written wasn’t just a record of what you’d communicated, it was a record of what you’d thought, and the principles were embedded in it all along.
That article was about you: your writing, your principles, your thinking. But this month, I want to zoom out to consider the team you run and the company you’re in. What happens when the same idea operates at the level of an entire organisation? Not just one leader extracting their principles, but a company where writing is the default medium for thinking, deciding, and communicating.
There’s a distinction here that matters. Most organisations have a documentation culture, where writing is an afterthought, something you do after the decision has been made to record it for posterity. The best organisations have a writing culture, where writing is the default medium for making decisions, not just recording them.
Here’s a test you can apply right now. Does your team write to communicate decisions that have already been made, or to make the decisions? If it’s the former, you have documentation culture. If it’s the latter, you have writing culture. Keep that distinction in mind as we proceed through the article.
This is a deep dive for subscribers, so we’re going to go long on this one. Here’s what we’ll cover:
The gold standard. Organisations that run on writing, and what we can learn from how they do it.
Three patterns that matter. The cross-cutting themes that separate writing cultures from documentation cultures.
What this looks like in practice. My own weekly write-up, and what it’s taught me about writing as a leadership tool.
Starting small. What you can try on Monday, whether you’re an individual contributor or running a department.
Why writing cultures fail. The patterns that kill writing cultures before they take root.
Writing in the age of AI. Why AI makes writing more important, not less, and the new failure mode it introduces.
If you’d like to explore related themes, here are some articles from the archive:
The importance of writing makes the individual-level case for writing as a leadership skill. This article zooms out to the organisational level.
A weekly mind meld covers sharing your thinking at scale through writing.
Going direct explores communicating without intermediaries, which writing cultures enable by default.
The spectrum of synchronousness looks at where writing sits on the async spectrum.
Let’s get into it.
The gold standard: organisations that run on writing
Let’s start with the absolute ceiling. Who does it best? These aren’t examples to copy wholesale. They’re illustrations of what becomes possible when writing is the default medium for thinking and deciding, not just for recording what’s already been agreed.
It’s worth noting that these are snapshots of particular eras in time: Bezos has since stepped back from Amazon’s day-to-day, Singleton has moved on from Stripe, and HashiCorp has been acquired by IBM. The people and the org charts have changed since the cultures were documented, but the practices they established are what matter here.
Berkshire Hathaway: writing as the vehicle for trust
The infamous Berkshire Hathaway has a portfolio that spans insurance, railways, energy, retail, and manufacturing, with fully owned subsidiaries like GEICO, BNSF, and Duracell operating alongside major stakes in companies like Apple and Coca-Cola. At the time of writing, the company is valued at over a trillion dollars.
The entire operation is overseen by around 27 people at headquarters in Omaha. Yes, just 27!
Far from being a quirk, that ratio is the result of a decentralisation philosophy where subsidiary CEOs operate with almost total autonomy, what Munger called “delegation just short of abdication.” The lean structure comes from the culture, not the writing. But, importantly, writing is the vehicle through which that culture transmits and sustains itself, most visibly through Buffett’s shareholder letters, which have become some of the most studied examples of business writing in the world.
Buffett’s approach to writing is famously simple: he imagines he’s writing to Doris and Bertie. As he put it in the preface to the SEC’s Plain English Handbook, “I pretend that I’m talking to my sisters. I assume they’re smart but not experts.”
The result is decades of shareholder letters, written clearly enough that anyone can follow the reasoning, and substantive enough that business schools still teach from them. His Owner’s Manual, first published in 1996, laid out the principles for how Berkshire operates and was reattached to every subsequent annual report for over twenty years: a constitutional document, not just an executive summary.
Bezos did something similar at Amazon with his 1997 shareholder letter, which we’ll come to shortly.
Then there’s the biennial memo Buffett sent to subsidiary CEOs for over 25 years, a single page with two mandates: guard the firm’s reputation, and plan your succession. “We can afford to lose money, even a lot of money. But we can’t afford to lose reputation, even a shred of reputation.”
That’s the entire management directive for a trillion-dollar portfolio, and it worked because it was written clearly enough that nobody could misunderstand it, and repeated often enough that nobody could forget it.
Simple and effective.
Amazon: writing as a thinking tool
On June 9th, 2004, Jeff Bezos sent an email to Amazon’s S-Team, the company’s most senior leadership group, that banned PowerPoint.
The reasoning was blunt. “The narrative structure of a good memo forces better thought.” In its place, he introduced the six-page memo: a narratively structured document that had to be written in full sentences, with real verbs, real topic sentences, and real arguments.
As he later told Fortune, “full sentences are harder to write. They have verbs. The paragraphs have topic sentences. There is no way to write a six-page, narratively structured memo and not have clear thinking.”
The practice that made the memo culture stick was what Bezos called “study hall“: twenty to thirty minutes of silent reading at the start of every meeting, with everyone working through the memo before any discussion begins. “If we don’t,” Bezos explained, “the executives, like high school kids, will try to bluff their way through.”
The idea has since spread well beyond Amazon: Square adopted silent meetings after Alyssa Henry brought the practice from AWS, and Dropbox’s Drew Houston introduced memo-first meetings across the company.
I’ve been using this approach myself for complex documents going all the way back to my time at Brandwatch, and it’s remarkable how much sharper the discussion becomes (and how there is less pressure to read everything in a rush before the meeting).
In his 2017 shareholder letter, Bezos went further, deliberately slowing down strategic work: great memos aren’t written in a day. “The 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.” This is System 2 thinking by design: forcing the slow, effortful cognition that slides and real-time brainstorming actively work against.
Perhaps the most striking example of Amazon’s writing culture is the PR-FAQ, described in detail in Colin Bryar and Bill Carr’s Working Backwards. Before any new product gets built, someone writes a press release as if it’s already launched, followed by a set of frequently asked questions that stress-test the idea from every angle.
AWS spent months writing and rewriting PR-FAQs before a single line of code was written for S3 and EC2. Most PR-FAQs don’t get approved, and that’s the point. The writing process kills bad ideas cheaply, before engineering resources are committed.
I saw a similar dynamic while I was at Shopify, where every project goes through GSD, an internal review process that requires teams to make a convincing written case for the problem they’re solving and how they’ll solve it. As one of the OK1 reviewers, I’d review these proposals asynchronously alongside directors from product and UX before any work could proceed. Often, project and transition proposals were deeply researched, typically with supporting videos, prototypes, and data analysis.
Berkshire, Amazon, and Shopify are very different companies, but the same underlying dynamic is at work in all of them: writing isn’t an afterthought, it’s the main event. The question is what makes that work. Strip away the brand names and the specific formats, and three patterns emerge.
Three patterns that matter
So, what ties these examples together? Berkshire, Amazon, and Shopify are wildly different companies, but the writing cultures that underpin them share three common features. Before we move to what this looks like in practice, let’s name them.
Writing is the default, not the exception
The companies that get this right don’t add writing on top of their existing meeting culture: writing replaces the meeting. Amazon’s six-pager doesn’t supplement the presentation, it eliminates it. At Basecamp, every product bet begins with a written pitch evaluated at the “betting table”, never a slide deck or a verbal proposal. As Jason Fried has put it, “The pitch is always written. It’s never performed.”
Stripe operates the same way: as Dave Nunez described it, “From leadership on down, we default to writing.” At HashiCorp, executives execute their primary responsibilities through writing, and the company’s PRDs and RFCs are the mechanism through which decisions are made, not a record of decisions that happened somewhere else.
The distinction matters because it changes the incentive. If writing is an addition to your existing workflow, it’s overhead: one more thing to do, on top of the meetings you’re already attending. If writing is a replacement, it’s a trade. You get time back.
Basecamp’s internal communications guide captures this neatly. “Writing solidifies, chat dissolves.“ The shift isn’t “write more.” It’s “stop doing the other thing.”
This is the hardest pattern to adopt, because it requires you to remove something people are comfortable with. Writing a memo is easy to mandate, but try cancelling the meeting it replaces and you’ll quickly find where the real resistance lives.
Leaders write first
Every writing culture in the previous section has the same tell: the most senior people are the most prolific writers. Buffett has written the shareholder letter himself, up until his recent retirement, for over four decades. Bezos didn’t just mandate the six-pager, he reads every memo carefully and asks detailed questions about the content. Patrick Collison sends internal emails with footnotes.
David Singleton, Stripe’s former CTO, published at least one internal blog post per month, arguing that “investing extra time to communicate an idea through clear, precise writing delivers outsized results because vastly more people consume the writing than produce it.”
Singleton’s point reveals the key leverage of writing. It’s asymmetric. A meeting reaches whoever is in the room, but a document reaches everyone who needs the information, now and in the future.
When leaders write, they’re making a rational investment: trading a larger upfront cost for vastly greater reach. When they don’t, they’re optimising for their own convenience at the expense of their organisation’s information flow.
This is why mandating a writing culture without modelling it never works. If the VP of Engineering communicates through Slack messages and ad-hoc conversations, that’s the culture, regardless of what the process documentation says. The modelling is the mandate. We’ll return to this when we get to failure modes, because it’s the most common way writing cultures die.
Writing compounds
Here’s the pattern that makes writing culture fundamentally different from meeting culture. Meetings evaporate, but documents appreciate. A meeting is a one-time event that benefits whoever was in the room. A document is an artefact that compounds over time, gaining value as more people read it and more decisions reference it.
Buffett’s 1997 shareholder letter is still relevant twenty-nine years later: not as a historical curiosity, but as a living document that new Berkshire shareholders read to understand how the company operates. Amazon’s PR-FAQs become the canonical record of why something was built, a resource that product teams reference years after launch.
GitLab’s handbook, which would run to over 2,000 pages if printed, serves as the institutional memory for a company spanning over 65 countries. GitLab enforces this by keeping only 90 days of Slack history: if it matters, it belongs in the handbook, not in a chat thread that will disappear.
Basecamp’s internal guide makes the case plainly. “Speaking only helps who’s in the room, writing helps everyone. This includes people who couldn’t make it, or future employees who join years from now.“
Think of each document as a small deposit. A single memo doesn’t transform anything. But over time, the archive becomes institutional memory: searchable, referenceable, and available to people who haven’t been hired yet. Meetings don’t compound. Writing does.
These three patterns are visible at trillion-dollar companies, but they don’t require trillion-dollar scale. In the next section, I’ll ground them in something much smaller: a weekly write-up that takes thirty minutes and reaches an entire department.
What this looks like in practice
So, these patterns sound compelling when you’re talking about Amazon and Berkshire Hathaway, but do they actually work at normal scale? Here’s what happened when I tried.
