Lulu Cheng Meservy — a communications strategist who has worked with Substack, Anduril, and Activision Blizzard — explains how founders and companies can build audiences, shape narratives, and spread messages in a world where the old PR playbook is dead. The core idea is “go direct”: founders speaking personally and unfiltered to their audiences, rather than hiding behind corporate comms, press intermediaries, or generic messaging. She walks through a full comms framework built on message, medium, and messenger, with practical advice on crafting slogans, launching products, harvesting attention, writing apologies, and taking stands without alienating your base.
Go direct — what it means and why it matters now
Go direct means the founder speaks to the audience in the first person, without middlemen filtering the message through corporate PR language.
It does not mean doing everything yourself forever, boycotting the press, or only tweeting.
It means the founder’s personality, vision, and motivations are present in every external communication — blog posts, events, regulator conversations, employee messaging.
The founder is the only person who can talk about the vision in the first person.
Why it matters now: opinions and information are no longer curated by a handful of TV channels or news outlets. A random anonymous account can spawn a meme that spreads faster than CNN. In this decentralized environment, if you don’t speak directly, people don’t know whether to trust you, what you stand for, or why they should want to work for you.
The energy of going direct: writing that comes from a real person has fire and conviction — even if it’s imperfect. Corporate speak is polished to the point of being dead and boring. Bad honest writing beats perfect sterile writing.
Engineering spontaneity: people now deliberately create moments that feel unscripted — filler words in speeches, car TikToks, leaving the camera button-press in videos — because spontaneity is so effective at building trust. The culture has shifted from “perfect on camera” to “the delta between who you are at the bar and who you are on camera should be zero.”
Ship-to-Yap ratio
Founders need a healthy ship-to-yap ratio — the balance between talking and actually building.
Elon tweets constantly, but the companies are shipping and growing. Palmer Luckey tweets a lot, but Anduril ships relentlessly.
The red flag is when a founder is tweeting all day but nothing substantive is coming out of the company — “shipping blogs, shipping tweets, but where’s the product?”
Palmer Luckey as a case study in deterrence: he has one of the highest “deterrence scores” of any founder — if he says he’ll do something, people believe he’ll do it, whether it’s achieving a goal or holding a grudge. This kind of conviction is strategically valuable.
Anduril’s “Don’t Work at Anduril” campaign: a masterclass in turning attention into recruiting. It wasn’t just a video — it was a months-long coordinated effort involving a website, a recruiting funnel, and S-tier people in comms, design, and strategy. The attention was converted into inbound job applicants, not just a fleeting viral moment.
Shaping a comms strategy — message, medium, messenger
Start with the business goal, not ego. What are you actually trying to accomplish?
Most commonly: recruiting (the war for talent determines company success), but also fundraising, regulatory approval, or enterprise sales.
The entire job of comms is to make the right people hold the beliefs that will trigger them to take the action you want.
Things in your direct control (salary, workplace quality) vs. things outside your control (whether people know you exist, want to work there, accept your offer). Comms addresses the latter.
How to craft your message
Message is the highest leverage thing to get right. If the message is bad, no amount of podcast appearances, press hits, or tweet formatting will save you. It’s like finding product-market fit — the message is the product.
Great messages don’t have to be novel. They take something people already feel but can’t articulate, and give it a name and shape.
“Founder mode” went viral because it gave founders permission to feel what they already felt — that being called a micromanager was actually just managing, and there was nothing wrong with them. The term provided catharsis.
“Go direct” itself was not coined from scratch — Brian Armstrong and others had used it, but Lulu reshaped it and attached her framework to it. You don’t need a new term; you need to reshape an existing one.
The message must attach to a “receptor” — something the audience is already thinking about and interested in. Don’t try to create a new obsession; plug into an existing one.
Example: if you want to recruit ML engineers, show up on Hacker News, Scott Alexander’s blog, or Dwarkesh’s podcast — not the New York Times or Joe Rogan. Smaller audience, but the right audience.
Stripe’s early strategy: go after developers specifically on Hacker News.
The candy coating analogy: your message is the medicine; the hook is the candy coating. You choose the flavor — the thing people already care about — so they’ll take the medicine. Without a candy coating, people won’t engage with an abrupt pitch.
Two approaches to timing the zeitgeist:
High conviction: watch the cultural direction closely, form a thesis, prepare heavily, and commit to a piece of intellectual territory before everyone else gets there. Play to where the market is going, not where it is today.
Dollar-cost averaging: put out constant iterations, most of which won’t hit. By definition, the flops are forgotten. The hits get all the traction and are what people remember. Lower risk than it feels.
Say what no one else will: the most powerful thing you can do is put form and shape to an energy that’s already in the culture but doesn’t have a name yet. “The writing is on the wall” — the future is already here, it’s just not evenly distributed (William Gibson).
What AI can’t do in writing: feel genuine emotion, speak from lived experience, or have conviction. AI can simulate emotion but can’t feel it. Emotion produces the “wrongness” that makes writing sharp — breaking grammar rules, being ungrammatical, being raw. AI tends to sand off the wrongness that is the point. Lean into emotion, personal experience, and empathy — these are your monopoly.
Speaking from experience gives you a monopoly: if you’re saying something millions of other people could say, there’s no reason anyone should buy that message from you (with their attention). But if you’re the only person with your exact experience, reactions, and voice, 100% of people interested in that thing have to come to you. Peter Thiel’s “competition is for losers” applies to messaging.
Caveat: the “I spent 487 hours learning this” format has become engagement-farming slop. Once templates go viral, they’re emulated by people farming engagement, and the format becomes disgusting.
The medium
Medium is how your message gets delivered — and it must show up where your target audience actually gets their information.
Cultural erogenous zones: the hot-button topics people are already obsessed with. Don’t try to create new interests; know where existing interests are and shape your message to be received by people with those interests.
How to identify cultural erogenous zones:
Who is talking about it — what status zone and intellectual “pocket” they occupy. Saturating one pocket can be enough.
What they think and how strongly — resentment and jealousy show up loudly and publicly; cringe and embarrassment are silent. The worst kind of bad is when everyone’s embarrassed for you but no one says anything. Read what’s not in the room.
Trajectory — “Founder mode” went from cool to cringe in under 36 hours. Now it’s used ironically. Know whether something is burning out or sustaining.
The power of the right messenger
The messenger is the person with authority to say the thing. Words in a vacuum don’t mean much — credibility and authority give them weight.
Different messages require different messengers:
Vision and mission: must be the founder, because only they can say it in the first person.
“This is a great place to work”: should come from employees, not the CEO. A janitor saying the CEO greets them and asks about their children is infinitely more powerful than the CEO saying “I greet everyone.”
Market size and revenue potential: should come from investors or commentators, not the founder (distasteful if the founder says it).
Product quality: should come from users, not the company.
The mistake: using messengers whose opinions are discounted to nearly zero — paid spokespeople, generic PR pitches. Journalists hate getting hundreds of emails a day from people paid to send them. The messenger has no credibility, no passion, and everyone knows they’re being paid. You’ve neutralized all possible impact.
Context matters enormously: Matthew McConaughey in a Lincoln commercial is a forgettable shill. Matthew McConaughey on the UT football sidelines hitting the drum is beloved — because he’s there out of genuine love. Same person, two contexts, completely different reception.
Fake user-generated content (brands paying people or using avatars to simulate organic fans) is a one-shot bullet. Once people realize what’s happening, it’s permanently destroyed. You get one chance to fool people, and then they’ll never let you do again.
Sometimes the most powerful messenger is the most ordinary person: “Jane who lives down the street” talking about how things used to be — we trust her because she’s normal and real.
How to ask for amplification without being cringe
The problem with “can you share this?”: favors are not a renewable resource. You burn more relationship capital than you think. The recipient either (1) ignores it and feels guilty, (2) does it resentfully, or (3) genuinely wants to and does it happily. You only want category 3.
How to frame the request: don’t make it a CTA or an obligation. Instead, let them know what you’re doing because it matches something they’re already excited about or have been saying. “I thought this would be interesting for you” — no ask attached. If they’re in category 3, they’ll share of their own volition. If they’re in category 1 or 2, they have an excuse not to and don’t feel bad.
Two good comms tests:
Could you say what you’re about to say without a call to action asking for money in the next sentence? If not, it’s purely self-interested.
If your company didn’t exist, what could you say that would make your target audience feel understood? Start by relating to their pain, struggle, and hopes — then link back to the company later.
Taking a stand without alienating your base
Companies should be opinionated and take stands — if you stand for nothing, you’re a fungible commodity. People choose products that reflect the person they want to be.
But don’t divide your own people. The line in the sand should be “gerrymandered” to fit your people inside it.
Bad line: Republican vs. Democrat — splits your employees and users in half.
Good line: “We believe in X, and if you don’t, that’s fine — you weren’t going to help us anyway.” Example: Palantir and Anduril draw the line as “Western values and American greatness vs. those who oppose it.” Anyone outside that line was never going to work there, so there’s no cost to alienating them. In fact, it convinces your people that you’re real.
Another good line: crypto vs. non-crypto, or economic freedom vs. suppression — unites your base rather than dividing it.
Apple’s shift: the “Think Different” era made users feel understood — Apple was the tool for creators and outsiders. Now Apple feels like it’s pushing features to serve its own metrics (Genmoji ads in Settings), shifting from “they serve us” to “they ask us to serve them.”
How to launch your product
Three things to know about launches:
It’s incredibly noisy — launches happen multiple times a day now, and breaking through is very hard.
Launches are not the be-all and end-all. Nobody remembers the exact day Facebook or Airbnb launched. It’s okay if yours isn’t perfect.
Turn attention into something — a recruiting pipeline, sales pipeline, investor interest. Don’t let it evaporate as an ephemeral moment.
The mistake: jumping onto trends. Once manifestos went viral, everyone wrote manifestos. Once sizzle reels went viral, every launch was a sizzle reel. Once something works, everyone does it, and it becomes tiresome. You have to constantly find the next thing.
Trend vs. cultural erogenous zone: they can overlap. A cultural erogenous zone is something people genuinely care about right now. A trend is a format that’s being overdone. The key is to tap into the underlying interest, not copy the surface format.
Stop making narcissistic announcements: the standard outline (summary of news, what the company does, the problem, the solution, why it matters, CTA) is solid. Where people go wrong is navel-gazing — “I’ve always wanted to do this, congratulations to me” — without telling the reader why any of it matters to them.
Pressure equals force over area — focus your message
To break through, concentrate force on a small area. Don’t talk to 5,000 people hoping the right 5 find you. Talk to 5 people about the one thing they care about most.
Steve Jobs’ Stanford commencement speech: most people remember one thing — “Stay hungry. Stay foolish.” If the best speech of Jobs’ life is remembered for one line, don’t expect people to remember 10 things from your average blog post. Choose: do you want people to remember one thing or zero things?
What slogans work
Every great movement has a slogan. Slogans make something feel ubiquitous and inevitable.
The key: say the same crisp words over and over. If you hear the same message 100 times in different forms, it’s background noise. If you hear one phrase 100 times, it’s pressure — it cuts through.
But don’t repeat entire paragraphs — that’s robotic and scripted. Repeat the core phrase (“Think Different,” “Move Fast and Break Things,” “It’s Always Day One”) while the stories and examples around it change.
The Far Side cartoon analogy: a dog hears “blah blah blah REX blah blah blah REX” — that’s how your audience hears you. Choose the one word or phrase that cuts through, and drive that.
Converting attention into business outcomes
Attention is a means to an end, not the end itself. It must be converted into motion — recruits, revenue, fundraising. Otherwise it’s a dopamine treadmill and a sinkhole for effort.
The funnel must line up with something people already want to do. If you want recruiting conversions, understand what your target hires want (competitive salary, mission, restoring Western military dominance) and offer or speak to those things through the founder’s authority, third-party vouching, and a resonant mission statement.
How to write an apology — the CrowdStrike rewrite
The original CrowdStrike message was passive third-person: “CrowdStrike is actively working with customers impacted by a defect found in a single content update.” It sounds evasive — “whose defect? who put it there?” — and the founder is absent.
Lulu’s rewrite: “I’m the CEO of CrowdStrike. I’m devastated… I’m sorry… This came from an error that ultimately is my responsibility.” First-person, emotional, accountable.
Why it works: when a customer is upset, they need to feel the company is at least as upset as they are. If the company seems unbothered, the customer has to drag them down to their level before any real conversation can happen. If the company immediately meets them at that level — “sackcloth and ashes” — the customer often calms down and meets them back up.
The coat shop story: a clerk would fetch the boss, who would get even more upset than the customer about the defective coat. The customer would then say “it’s not that bad” and try to calm the boss down. The boss chose the path that brought them to the customer’s level immediately, rather than forcing the customer to drag them down.
