Rostra

Reuse a familiar term instead of coining a new one. When Lulu Cheng Meservey was starting Rostra and considering a new term for her approach, Brian Armstrong told her to just use "go direct." She took the existing phrase, reshaped it and attached her own ideas to it, and it became the shorthand for everything she teaches.

Borrowing a phrase people already know gets you familiarity for free. Armstrong's advice: "you don't need a new term."

How I Write▶ 21:30

The founder's own voice should reach the audience without middlemen. Publicists and journalists once controlled company narratives. Opinion is now decentralized, and a meme from an anonymous account can spread faster than a CNN story, so founders have to speak to the public themselves.

Going direct does not mean doing everything yourself forever or boycotting the press. It means some of the founder is in every channel: blog posts carry the founder's ideas, the founder hosts events (some host dinners at home), and keeps texting-level relationships with regulators instead of leaving them to lobbyists. "Without the founder, everything about your company is just flatter."

How I Write

Honest and imperfect beats polished. Audiences now want almost no gap between your public and private self, someone who sounds the same on camera as at the bar. Broadcast training aimed at perfect delivery is now the wrong target.

The Obamas add filler words to rehearsed speeches, creators leave the arm pressing record in the shot, and Congress members film vertical videos with handheld mini mics. Faked authenticity shows when done badly, so the best course is to be yourself. "I would rather see writing that is suboptimal, that has personality, that has intent and has conviction than writing that is like textbook correct."

How I Write

A few public missteps are worth the leverage of a direct platform. People often cite Elon Musk's mistakes as a case against going direct. Meservey argues his platform lets him fact-check hostile voices himself, and anyone would take that deal for the few mistakes it cost.

How I Write

Find the message like you find product-market fit. The message is the highest-leverage part: a weak one wastes every podcast, press hit and tweet spent delivering it, like a route selling encyclopedias nobody wants. Test it in conversation until it comes down to a few exact words.

Peter Thiel hosts dinners and works through his messaging the way comedians develop jokes.

How I Write

Name a feeling people already have and tell them it's legitimate. "Founder mode" spread fast because it gave shape to something founders already felt. The strongest message says what everyone in a field thinks and nobody says, turning energy with nowhere to go into something solid.

It needn't be controversial. Themes also turn: "founder mode" went from cool to cringe in under 36 hours and is now used only ironically. Resentment is loud while cringe is quiet, so watch for usual supporters going silent.

How I Write

Wrap the message in a hook the audience already cares about. People aren't waiting to hear your message, and an abrupt jump to "my thing" usually gets a no. The message is the medicine and the hook is its candy coating, attached to a "receptor" already on the audience's mind.

Her own example: interest in AI is the receptor, and the bridge is that you wonder how to use AI and writing is one of the most important things you do. The coating is the way in, not a trick to hide the message; a recruiting pitch should be built around what candidates already follow.

How I Write

Pick the medium and messenger the right audience trusts. A smaller audience of the right people beats a huge wrong one; Stripe went after developers on Hacker News. A janitor praising the CEO carries more weight than the CEO praising himself.

Dwarkesh's podcast or a Scott Alexander or Tyler Cowen post reach the right engineers, though Dwarkesh is harder to get onto than Rogan.

How I Write

Claim an idea by timing it or by iterating constantly. Timing means forming a thesis and playing to where an idea is going, early but not irrelevant. Iterating means putting out many posts, keeping what hits and refining from feedback.

A flop is lower risk than it feels because, by definition, nobody saw it.

How I Write

Work backward from a business goal to the beliefs that trigger action. Attention only counts when it becomes recruits, sales or funding. Start with what the business needs, then work out what people must believe; recruiting is the best default goal because the war for talent decides a company's success.

Anduril's recruiting campaign took months and included a website and a funnel, so the attention became hires. It also said plainly who it didn't want with a "don't work at Anduril" stance. Most startups make a video in days and get a few hours of attention, which is potential energy that never becomes kinetic.

How I Write

A founder's shipping must outweigh their posting. Founders who ship constantly can post constantly; Musk tweets heavily yet nobody thinks his companies are neglected. A bad ship-to-yap ratio, blogs and tweets while people ask where the product is, is a red flag.

At Anduril, Palmer Luckey tweets a lot while Brian Schimpf just ships, and Luckey is credible because people know he does what he says.

How I Write

Don't ask for shares; give people something they already want to say. A favor request leaves three outcomes: they resent you, comply and feel icky, or were happy to share anyway. Only the last is worth having, so share it as an FYI that fits what they care about and let willing people spread it.

How I Write