There is a graveyard of products that were better than the thing that beat them. Nobody visits it, because nobody heard of them in the first place.
That sentence is the whole argument, but founders resist it. When you build something, you want to believe quality is the scoreboard. Make the best thing and the world will notice. It is a comforting belief, because quality is the part you control while you sit at your desk.
It is also mostly wrong. The best thing rarely wins. The best-known good thing does.
The myth of the self-evident product
Picture two apps that do the same job. One is slightly better designed, slightly faster, slightly more thoughtful. The other is decent, and it has a way of reaching people: a big partner, a viral loop, a founder whom fifty thousand people already read.
In almost every market, the second one wins. Not because users are foolish, but because users cannot choose something they have never seen. Quality only matters after attention has done its job.
This is why Peter Thiel's Zero to One spends a whole chapter arguing that engineers chronically underrate distribution. People who build things tend to see sales and marketing as a tax on real work, something slightly embarrassing that happens after the product is done. So they treat it as an afterthought, and the afterthought decides the outcome.
The more useful mental model is that distribution is a feature. It belongs on the same whiteboard as the data model and the onboarding flow. How will the thousandth user hear about this? If you cannot answer that, you do not have a product yet. You have a nice object.
A blog that changed a career
The clearest example I know does not involve a startup at all. It involves a young researcher with a blog.
In the mid-2010s, Chris Olah began publishing long, careful posts about how neural networks work. One of them, "Understanding LSTM Networks," explained a type of model that confused nearly everyone who tried to learn it. Papers described it with walls of equations. Olah described it with clean diagrams that showed information flowing through the network step by step, with each piece introduced only when you needed it.
The post became one of the most widely shared explanations in machine learning. For years, if you asked someone how they finally understood LSTMs, there was a good chance the answer was that blog. Other posts, on topics like how neural networks bend and fold data, had a similar effect. Students learned from them. Researchers sent them to colleagues. Professors pointed classes to them.
Now look at what followed. Olah did not come up through the usual path of a famous PhD programme. He went on to do research at Google Brain, then led interpretability work at OpenAI, and in 2021 he was one of the co-founders of Anthropic. Along the way he also helped start Distill, a journal built on the idea that explaining research clearly is itself a form of research.
It would be unfair to say a blog did all that. The work was excellent, and that is the foundation. But excellent work is common in research. What was rare was that thousands of people already knew his name and trusted his thinking, because he had spent years being genuinely useful to them in public.
His writing was not marketing for his research. It was the research, made reachable.
That is distribution, in its purest form. Nobody bought an ad. Nobody ran a growth campaign. He explained hard things more clearly than anyone else, and the explanations travelled on their own.
Why writing works so well
Writing is the most underrated distribution channel available to anyone who builds things, for a few specific reasons.
It compounds. A good post written five years ago is still finding readers today. A conference talk ends when you walk off the stage. An ad stops working the moment you stop paying. An essay that genuinely teaches something can keep working for a decade.
It filters for the right people. When you write clearly about a hard problem, the people who read to the end are disproportionately the people who care about that problem. Those are your future users, collaborators, hires and investors. Writing is a sorting machine for the people you most want to meet.
It proves the thing it describes. Saying you understand something is cheap. Explaining it so well that a stranger understands it too is proof. A clear essay is a portfolio piece and an argument at the same time.
It is cheap and permissionless. You need no budget, no gatekeeper and no audience to start. You need a problem you understand and the patience to explain it properly.
Paul Graham is the other famous example. His essays on startups were read by a generation of founders before and during the early years of Y Combinator. Whatever else you think about the accelerator, it is hard to imagine its early pull without that body of writing. Many founders applied because they had been reading him for years and felt they already knew how he thought.
What this means if you are building
If distribution is part of the product, then a few practical things follow.
First, decide how people will find you before you finish building. Not as a marketing plan in a slide deck, but as a real answer. Is it search? Word of mouth inside one specific community? Partnerships? A founder with a following? Each one changes what you should build. A product that spreads by word of mouth needs a moment worth telling a friend about. A product that spreads through search needs to answer a question people are already typing.
Second, pick channels where you have an unfair advantage. In the Gulf, for example, a lot of business still moves through relationships and trust built over years. A founder who already belongs to the right community can reach customers in a week that an outsider could not reach in a year. That is a distribution asset, and it is worth being honest about whether you have it.
Third, start early, because distribution is slow. An audience, a reputation, a set of partners: all of these take time to build and cannot be bought quickly once you need them. The founders who seem to launch out of nowhere usually spent years quietly building the channel before the product existed.
Writing as a personal distribution system
For individuals, the lesson is even simpler.
Most people treat their best thinking as private. They solve a hard problem at work, learn something surprising, form a sharp opinion about where a field is going, and then tell nobody. The knowledge stays locked in one head.
If you write it down, clearly, in public, the same knowledge starts working for you while you sleep. Someone you have never met reads it, learns from it, and remembers your name. Do that consistently for a few years and you have built something that no CV can match: a body of evidence about how you think.
There are a few rules that make this work.
- Teach, do not perform. The writing that travels is the writing that leaves the reader genuinely more capable. Clever is nice. Useful is what gets shared.
- Go deep on something narrow. Olah did not write about everything. He wrote about how neural networks work, again and again, better than anyone else. Depth makes you the default link.
- Respect the reader's time. Cut every sentence that does not earn its place. Clarity is a form of generosity, and readers feel it.
- Be patient. The first posts will reach almost nobody. That is normal. Compounding looks like nothing for a long time.
The quiet advantage
There is something slightly unfair about all of this, and it is worth naming. Two people can have equal talent, and one of them ends up with far more opportunities simply because more people know what they can do. Nobody chose that outcome on purpose; attention simply flows toward what it can see.
The good news is that this particular unfairness is one you can do something about. You cannot always choose your talent, your starting point or your luck. You can choose to explain what you know, clearly, where people can find it.
The world does not reward the best work it never sees. It rewards the good work that reaches it.
Build the thing. Then build the road to it. And if you can only afford one road, make it your own words.
Further reading
- Zero to One — Peter Thiel with Blake Masters
- Hackers and Painters — Paul Graham
- Traction — Gabriel Weinberg and Justin Mares



