How to cold email a startup founder

Does cold outreach to founders actually work?

Yes, and it is the main thing separating the people who get interviews from the people in the pile.

Founders are not protected by gatekeepers at this size. There is no recruiting team, no switchboard, no assistant. The CEO's LinkedIn is public and they check it themselves.

The messages below took four minutes each to write. Several of them produced a call within a day.

LinkedIn, email, or something else?

Start with LinkedIn, fall back to email. The connection request is doing useful work before you have written anything.

A founder only accepts your request after looking at your profile. So an accepted request means they have already looked at you and decided you were worth a moment. Psychologically they are half-committed to a conversation before your first message arrives. That is why your profile matters more than your CV at this stage.

If the request is not accepted, email. Not a demo booking or a sales address used as a back door, just their actual address, which is often on the site or findable from their public writing.

Two channels worth knowing about. In San Francisco, X carries real weight if you have any following there. And for engineering or product roles, if the company has anything open source, opening a pull request that adds something genuinely useful is a stronger first message than any email, because they can see your work instead of reading a description of it.

How long should the message be?

Three or four sentences. It should be fully readable on a phone without tapping anything.

Every message below is under sixty words. They contain who you are, why this company, and a clear ask. Nothing else fits and nothing else is needed.

The ask should be small. A call, an introduction, a question. Not a job.

What should the message actually say?

Name the specific reason it is this company, and end with one small, easy ask.

Five messages that worked, with the situation each was written for. Every one of these led to a call.

The pattern across all five: a specific reason it is them, an honest statement of what you want, and an ask that takes one word to answer. None of them mention passion, none of them list qualifications, none of them are longer than a text message.

What does a bad cold message look like?

Generic enthusiasm with nothing specific in it. It reads as AI-written whether or not it was.

This one got no reply, and deserved none.

Resonates, vision, mission, contribute. Every word of it could have been sent to any company on earth, which is exactly what makes it invisible. A founder reading it learns nothing about you and nothing about why you picked them.

The follow-up three days later did work, because it had actual content in it: that the first application went through LinkedIn Easy Apply with no space for any motivation, and that a second one had gone in through the YC platform where there was. Specific, and it gave a reason for the message to exist.

Does building something for them first work, or does it look try-hard?

It works, as long as it is evidence of what you can do rather than a gift.

The distinction is whether the thing shows your work or asks for their attention. A pull request on their repo shows your work. A forty-slide unsolicited strategy deck asks for their attention.

• Engineering: a GitHub profile, or a PR on something of theirs that adds something real
• Product: a portfolio, or a teardown of something in their product with a specific suggestion
• Go-to-market and growth: your own numbers, an account you grew, content you shipped, an agency you ran

A Loom under a minute also works well in place of a long email. Who you are, what you have done, why them. It is harder to ignore than text and it shows how you communicate, which is most of the job at a small company.

How often should you follow up?

Every week or two, three or four times, each one short and with no pressure in it.

Not every three days. That is spam, and it changes how they think about you in a way you cannot undo.

Here is a real sequence with a YC-backed company in San Francisco, spread over two months of near-silence. Four messages. Every one of them got a warm reply.

The reply came eleven days later. She apologised for the delay, said they were genuinely interested in working together, and that the open question was timing and workload rather than whether they wanted him. That is what silence usually is.

Six days after that she messaged first, to say she would be in Belgium and to set up a coffee. Two months of stalling ended because he had stayed visible without ever applying pressure.

Look at what each message does. It is short. It gives a reason for existing right now, like the end of the month or the holidays being over. It restates interest once. And it ends by handing control back to them. No guilt, no urgency, nothing that needs a defensive answer.

Why do founders go quiet in the middle of a process?

Almost never because they have decided against you. Usually timing, workload, or a decision they have not made yet.

This is the single most useful thing to understand about startup hiring. The founder who has stopped replying is head down shipping something, or waiting on a round to close, or has not worked out whether they can afford the role this quarter.

Candidates read silence as rejection and stop, which is the one response guaranteed to end it. A rejection is a rejection. Silence is a queue.

There is a second thing to do during a quiet stretch that almost nobody does: build a relationship with someone else at the company. In the example above, the third founder had not been met yet, so he reached out separately and had a conversation with him during the silence. The CEO's next message opened by referencing it. He was no longer a name in an inbox, he was someone her cofounder had spoken to.

What do you do when it stalls for weeks?

Ask a direct question. Two months of polite check-ins produce polite answers, one blunt question produces the truth.

After the coffee, and after weeks of updates that never quite became a decision, the message that finally moved things was five words long: where's your head at right now.

That got a real answer. They expected to be scaling from Q2 but the timeline was uncertain, she wanted to work with someone as hungry as she was, and the actual blockers were location and a visa. None of that had come out across two months of professional back-and-forth.

The other lever is honest competition. If you are genuinely talking to other companies, say so, plainly and without threatening anything. It removes any sense that you are waiting around and it makes their timeline their problem rather than yours. Only use it if it is true, because founders can tell, and a bluff that gets called is worse than saying nothing.

What if you have been waiting weeks on a company you want?

Then send the message. The silence is not happening to you, it is a message you have not sent.

Seven weeks of waiting is seven weeks of choosing not to write two lines. In the example above, every single check-in got a warm reply, and the process only ended when he decided it should. He was asked to fly out to San Francisco for a trial after four conversations. He did not go, because by then his own project had started working.

Nothing bad happens when you send a short, friendly check-in to someone who has already shown interest. The worst realistic outcome is that they tell you the role is not happening, which is information you wanted anyway and which frees you to spend the effort somewhere else.