The Best Marketing Essay Ever Written Has Nothing to Do with Marketing

Four ideas from "Do Things That Don't Scale" that every marketer should know.

The best marketing I’ve ever done for startups felt nothing like “marketing” at the time. It looked more like customer service. Or consulting. Or just spending too much time on something that clearly wouldn’t scale.

Paul Graham explained why over a decade ago.

His essay “Do Things That Don’t Scale” is aimed at founders, but the more I reread it, the more I think it’s one of the best pieces ever written about early-stage marketing. It doesn’t use the word “marketing” much. It doesn’t need to.

Four ideas from the essay have stuck with me over the years and helped me build marketing that works:

1. Respect the Newborn Stage

Graham says people look at early startups like a newborn baby. They see this tiny creature and conclude it could never accomplish anything.

I see the same thing happen with marketing channels.

The first version of your best system always looks embarrassingly small. A handful of users. A few conversations. Numbers that wouldn’t survive a slide deck. And most teams kill it right there, before compounding can even start because someone looks at the early numbers, compares them to where they need to be, and pulls the plug. They move on to the next tactic. The next channel. The next idea that promises faster results.

The ones that survived that phase did so because someone had the patience to keep going when the numbers were small and the progress felt slow.

2. Build for a Person, Not a Persona

This is the one that changed how I think about positioning.

Graham’s advice is simple. Do not build for a broad demographic. Pick one user. Act like a consultant building just for them. If you get it right for one real human, you’ll usually find you’ve built something a lot of people want.

Marketing works the same way.

When you write for a persona, you end up with messaging that sounds like it was written by a committee. It checks all the boxes and moves no one. But when you write for a specific person you’ve actually talked to, someone whose frustration you remember, whose words you can hear in your head, the messaging starts to land.

The interesting thing is that the more specific you get, the more universal the work becomes. One person’s real problem, described in their real words, resonates more broadly than a carefully constructed composite ever will. The teams that talk to one customer deeply always outperform the teams that survey a thousand customers broadly.

3. Make the Fire Hot Before Adding More Logs

This one doesn’t get talked about enough.

Graham points out that Facebook didn’t launch to the world. It launched at Harvard. Just Harvard. It got critical mass there, then expanded to other Ivy League schools, then to other colleges. By the time it opened up broadly, it had already built real demand.

Graham describes this as keeping a fire under control to make it really hot before adding more logs.

The instinct in marketing is almost always the opposite. Go broad. Go fast. Cast the widest net you can. But the companies I’ve seen build the most durable growth did something different. They picked a small market, owned it completely, and then expanded from a position of strength.

In fintech, I’ve seen this firsthand. The startups that try to serve everyone on day one end up resonating with no one. The ones that pick a specific community, a specific use case, a specific problem and go deep on it, those are the ones that build real traction. The broader market comes later, and it comes easier because you’ve already proven something that works.

It takes discipline to stay narrow when everyone around you is telling you to grow faster. But narrow is how you build heat. And heat is what spreads.

4. Startups Are Powered Aircraft, Not Projectiles

This is the metaphor I come back to the most.

We often think of marketing in terms of a Big Launch. You build the campaign, you pick the moment, you fire it into the sky, and it either flies or it doesn’t.

Graham argues that startups don’t work that way. They’re not projectiles. They’re powered aircraft. You have to keep the engine running.

In my experience, the big rebrand or the massive campaign almost never provides the silver bullet. They can create a spike. But spikes fade. What actually builds a sustainable business is the steady, unglamorous work of showing up consistently, learning from what you see, and making improvements over time.

Growth is gradual. It has always been gradual. That is how sustainable businesses are built. The companies that accept this and build their marketing accordingly are the ones that tend to win.

What This Adds Up To

If there’s a thread running through these four ideas, it’s that the things that feel too small, too specific, or too manual to be “real marketing” are often the most important marketing work you’ll ever do.

Talking to one customer instead of surveying a thousand. Dominating a small market before chasing a big one. Nurturing a tiny channel instead of abandoning it. Committing to consistency instead of betting on a single moment.

None of it looks impressive on a slide. All of it compounds.

The essay is worth a read if you haven’t come across it. Worth a re-read if you already know about it. You can find it here.