Cold Emails from the James Hawkins Interview

I watched James Hawkins on this podcast and kept pausing on things he said -- not to understand PostHog better, but because he kept accidentally describing what he'd do differently if he were starting today. The interview is full of it.

Each email below uses something specific he said as a hook, then asks a version of the same underlying question: what would you build if you were a young engineer starting a software company from zero right now?

2. 1. Inbound-only business

PostHog does zero outbound at tens of millions ARR. No cold calls, no SDRs. The sales team exists to improve retention on deals that already came in. James frames this as a choice, not a constraint. It implies he picked a category and a motion where the product could pull customers by itself.

"We don't do outbound still at all. It's just cross sell and sales assist. We're a completely inbound business."

Draft

Subject: inbound-only from day one

Hello James,

You built PostHog as a fully inbound business -- zero outbound even at scale. If you were a young engineer starting fresh today, which category would you enter specifically because you knew you could make it inbound from day one?

Alexis (https://www.alexisbouchez.com)

3. 2. Six cycles to PMF

Before PostHog, James and Tim went through six build-validate cycles in six months. One of those was a technical debt tool that failed to convert at any price. The pattern he describes: build fast, talk to users, watch if they pay, abandon if not. Six times until something clicked.

"We had to go through like six cycles of doing this until we actually landed something that hit product market fit. Which took us like six months."

Draft

Subject: six cycles

Hello James,

You mentioned going through six build-validate cycles before PostHog hit PMF. If you were doing that again from scratch today, which domains would you be cycling through -- and what signal would make you stop on one?

Alexis (https://www.alexisbouchez.com)

4. 3. The CRM gap he named

James is building a CRM -- not to compete with Salesforce, but specifically for engineers doing their first six deals. "You're probably using Apple notes or a spreadsheet." He's building it because PostHog already has the user data and no good UI on top. But someone without PostHog could build just the CRM part and own that wedge.

"You're an engineer trying to do your first six deals. You're probably going to use Apple notes or a spreadsheet. Can we just make that PostHog instead?"

Draft

Subject: first six deals

Hello James,

You identified the "engineer doing their first six deals on Apple notes" problem and you're building it inside PostHog. If you were starting from scratch without the PostHog user base, would you still bet on that as a standalone product -- or does it only work as a bundle?

Alexis (https://www.alexisbouchez.com)

5. 4. The cloud move everyone avoids

PostHog almost didn't do cloud. Five competitors were already there. James thought they couldn't win. They shipped anyway, added a big free tier, growth jumped. He now describes it as obvious in retrospect -- "we just saw it happen." The lesson is about doing the thing that looks too competitive to be worth trying.

"We thought there's so much competition we were screwed. Like there's at least five alternatives doing a couple hundred million run rate. And we're like -- no, it turns out we can compete."

Draft

Subject: the move that looked too competitive

Hello James,

You almost skipped cloud because the competition looked too strong -- then shipped it and everything accelerated. If you were a young engineer today, what's the equivalent move that looks too competitive to try but probably isn't?

Alexis (https://www.alexisbouchez.com)

6. 5. Multi-product from day one

PostHog's strategy is to build every piece of software a software team needs. They went multi-product early and it compounded -- more customers, better employees, easier fundraising. James says the more ambitious they got, the easier things became. He seems surprised by this himself.

"The more ambitious we are, it kind of gets easier. More customers are more interested, employees are more interested, it's easier to raise money."

Draft

Subject: starting multi-product

Hello James,

PostHog went multi-product early and found things got easier the more ambitious you were. If you were a solo engineer starting fresh today, would you still try to go multi-product early -- or is that only something that works once you have the first product pulling customers?

Alexis (https://www.alexisbouchez.com)

7. 6. Label what's working

James's actual strategy, by his own description: do a bunch of stuff, label what's working, do more of that, do less of the rest. He applies this to marketing, to product, to hiring. It sounds simple but it requires having things actually working first. The question is what's working right now in dev tools for someone starting fresh.

"We've done a bunch of stuff. We've just labeled what's working and then done more of the stuff that's working and less the stuff that's not working."

Draft

Subject: what's working in dev tools right now

Hello James,

Your actual strategy -- "do stuff, label what's working, do more of that" -- makes sense once you have signals. If you were a young engineer starting from scratch today, what in dev tools is clearly working right now that you'd want to do more of?

Alexis (https://www.alexisbouchez.com)

8. 7. The AI coding tools debate

James mentions a debate PostHog had internally: who wins when AI can write code? Does the engineer role expand to cover everything else (sales, marketing, support), or does it shrink because code is no longer the moat? He can't tell. But someone starting a company today has to bet on one outcome.

"The debate we were having was -- with AI coding tools, who wins? Is it that the engineer is going to be expected to go further with everything else? Or is it that the product manager wins because engineering doesn't matter anymore?"

Draft

Subject: who wins with AI coding tools

Hello James,

You mentioned an internal PostHog debate about who wins as AI writes more code -- engineer expands to cover everything, or engineering stops being the moat. If you were starting a software company today, which outcome would you be betting on?

Alexis (https://www.alexisbouchez.com)

9. 8. Brand before product

James says if he were starting fresh today, he'd want a small social following first -- even a couple thousand is enough. His reasoning: it gives you real users for the first version, which speeds up the validation cycles. He spent six months cycling through ideas partly because he had no audience to quickly test with.

"If you can get a couple of thousand followers, I think you probably could get something off the ground with a few users to give you the learning you need."

Draft

Subject: audience before product

Hello James,

You said if starting fresh today you'd want a small following first -- even 2k is enough to validate faster. If you were building that audience from scratch as an engineer, what would you write about to attract the right people before you even had a product?

Alexis (https://www.alexisbouchez.com)

10. 9. What he'd bootstrap

James says if he had to restart tomorrow, he'd want creative freedom and less stress -- something smaller, something he could half-vibe-code. He explicitly doesn't want to go through existential crises about whether anyone cares. The question of what he'd actually build in that mode is left open.

"I'd find it fun to work on something with a bit of creative freedom and not stress about it as much. I'd be half vibe coding, trying to build something."

Draft

Subject: what you'd build if starting over

Hello James,

You said if you were restarting you'd want something smaller, more creative, less existential -- half vibe coding something. What would you actually build?

Alexis (https://www.alexisbouchez.com)

11. 10. The marketing stack gap

James admits PostHog has mostly ignored marketers as a persona. He knows the marketing tool stack well enough to list the gaps -- attribution, tag management, a GA alternative -- but hasn't built them yet. He's aware "marketers spend a lot of money on tools" and says PostHog may have been too slow here.

"It might be really stupid that we haven't done this already because word on the street is that marketing people will spend a lot of money on tools."

Draft

Subject: marketing tools for engineers

Hello James,

You said it might be stupid that PostHog hasn't gone harder on marketing tools yet. If you were starting a company today aimed at that gap -- marketing stack for technical teams -- where would you start?

Alexis (https://www.alexisbouchez.com)

12. 11. Engineer-led companies

At PostHog, engineers decide what to build. No product managers setting the roadmap. James argues this works because most companies apply Unilever-style management practices to early-stage contexts where they don't fit. His theory: people are more risk-averse than is rational, and the antidote is just hiring people you trust and then actually trusting them.

"You can trust developers to make good decisions at an individual level. The secret is: you can actually trust your own team. You should focus on building a team you can trust."

Draft

Subject: engineer-led product decisions

Hello James,

PostHog has no PMs -- engineers own the roadmap. If you were advising a young engineer starting their first company, would you structure it the same way from day one, or is that something that only works once you have the right team?

Alexis (https://www.alexisbouchez.com)

13. 12. 96% free users

96% of PostHog's 140k companies pay nothing. James is fine with this -- "if you're a hobbyist with no budget, there's no point trying to make money off you anyway." The free users say nice things online, which drives word of mouth for the paying ones. He'd structure it this way again. But it requires picking a space where this flywheel actually runs.

"If you're a hobbyist kind of user and you have no budget, there's no point trying to make money off you anyway. We want you to say nice stuff about us on the internet if you like our products."

Draft

Subject: 96% paying nothing

Hello James,

96% of PostHog users pay nothing and you're fine with it -- word of mouth from free users drives the paying ones. If you were starting fresh today, which other categories do you think that specific flywheel would work in?

Alexis (https://www.alexisbouchez.com)

14. The pattern

All of these ask the same thing: what would you do if you were me? They just use something he actually said as the hook, so it doesn't feel generic.

The hook matters because it proves you paid attention. "What would you build if starting over" with no context gets ignored. "You said you'd half-vibe-code something with more creative freedom -- what would that actually be?" is a real question that's fun to answer.

One question per email. Never two. The subject line is the hook, not the ask.