Founder-Led Hiring
Published on
Founder-led hiring means the founder personally finds, judges, pitches and closes every hire. Early on there is nobody else to do it, and the first people on the team set the culture, the bar and the product, so only the founder can say who fits. It starts with the first engineer and ends when recruiting becomes a team’s job. YC’s guide to hiring your first engineer tells founders to expect to spend at least a third of their time on those initial hires.[1] Most founders burn that third on a post, a DM and an inbox, and come out with nothing but noise.
This essay is my definition of the term and my argument for why the stage deserves a name. It comes from 16 years in and around early-stage engineering. I started as an engineer in 2010. Through the mid-2010s I was one of the top freelancers on the big platforms, with over $500,000 earned individually — which put me across the table from hundreds of founders who were trying to hire. I then ran a freelance agency, then a dev shop, and spent years as an internal hiring manager for hundreds of roles inside a large company. In 2023 I founded TeamEx as a marketplace for hiring engineers, and in October 2026 I narrowed it to one thing: founder-led, early-stage hiring. Across those seats I have interviewed more than 2,000 engineers and worked with hundreds of founders on their first hires. What follows is my take on founder-led hiring, built on that experience and on watching how successful founders handle it.
Founder-led hiring is Founder Mode applied to the team
In September 2024 Paul Graham published Founder Mode after a Brian Chesky talk at a YC event.[2] The essay’s target was one piece of advice every founder hears as the company grows: “Hire good people and give them room to do their jobs.” Graham’s point was that this is how to run a company you did not found, and that founders who followed it watched it go wrong.
Hiring has its own version of that advice. Write the job description, hand it to a recruiter, give them room. It arrives even earlier, often before the first engineer, and it fails for the same reason. The recruiter is matching a brief. The founder is choosing a person they will build with for years. Chesky now says he spends about three hours a day recruiting, and that Sam Altman was the one who changed his mind: “He said hiring should be 50% of your job.”[3] Altman had written the same thing a decade earlier, that a founder should spend “between a third and a half” of their time on hiring, and that “you can’t outsource this.”[4]
I think founders treat this stage as something to get through on the way to a real recruiting process. It is better understood as founder mode applied to the team. The founder keeps the judgment and hands off only the parts that do not need it. The hard part is that for the first few years, almost everything seems to need it.
It is low-volume and episodic, and that is the whole problem
A seed-stage startup hires a few people a year, one at a time, with months in between. Each hire is a large share of the company. At a two-founder startup, the third hire is a fifth of the team. The stakes are high and the volume is tiny, and that combination is what makes it hard.
The recruiting industry is built for the opposite. Recruiters manage about 10 open roles at a time and about 200 applications per hire. The tools are priced for that workload. LinkedIn Recruiter runs about $900 to $1,080 a month per seat on an annual contract,[5] and the median Greenhouse contract is $26,553 a year.[6] That stack, an applicant tracking system, sourcing seats, AI screening software and a recruiter to run it all, pays for itself if you have a steady hiring pipeline. At two hires a year it is dead weight, and most founders are right not to buy it.
I have hired at both ends of that range. With a recruiting team and a tool stack behind me, technical hiring was a pipeline I kept full. With two founders and a half-built product, every hire was a decision I had to get right, with nothing between me and the inbox. That second situation is the one nobody builds tools for.
If hiring is a war, the early-stage founder shows up without weapons, and the weapons for sale are built for armies. So the founder improvises.
How founders do it today, and where each path runs out
The improvised version looks the same at most startups. A “we’re hiring” post on X or LinkedIn. Cold DMs to engineers at companies the founder admires. A referral bounty. Hacker News’ monthly thread. When those stall, a freelance recruiter or an agency.
Every one of these can work, and each runs out in a predictable place. The post runs out first. A job ad draws 250 applicants or more, and at a startup the founder is the only person there to read them, without being equipped with tools to navigate the mess. DMs run out on time. When Harj Taggar said to spend at least a third of your time on hiring[1], I don’t think he meant desperate LinkedIn DMs that end in silence. Referrals run out when the network does, and a founder’s network always does. And the recruiter runs out on fit, because a role brief never captures how a two-person team works.
What most founders get wrong
Most of the mistakes come from what the founder believes hiring is. The tactics follow from that. These are the five I saw most often, first from the agency side, then from inside TeamEx.
Delegating it.
The first mistake is treating hiring as a task to get off the plate. A recruiter can find people and screen them. They cannot know what it is like to disagree with you when stakes are high, and they cannot make the pitch. Taggar’s guide is blunt about it: “Convincing engineers to join is one sales job you can’t delegate entirely to your cofounder.”[1] Delegate it to a stranger and it is gone.
I was that stranger. For years, founders sent me a brief and stepped back, and I did the job well by every measure a brief can hold. The people I sent could build what was asked. What the brief never contained was how the founder thinks, works, and communicates at work. Their principles. How they move and make decisions. I could not know that from a document, and the founders who got their best hires through me were the ones who stayed in the room and judged it themselves. I wrote up what that fit is made of separately, with three founders who hired through TeamEx and wanted three different kinds of engineer for the same title.
Rushing it.
The opposite mistake is giving it no time at all and hiring whoever appears first. Graham’s Do Things That Don’t Scale describes founders who “would rather sit at home writing code than go out and talk to a bunch of strangers.”[7] Hiring is the same unscalable thing, with the same temptation to skip it. Airbnb’s founders, by their first engineer’s own account, “had been looking to hire a first engineer for a long time” before Nick Grandy, and still ran him through a take-home project, an interview with each founder and a beer test.[8]
Describing instead of recognizing.
Most founders cannot write down who they want, and try anyway. The job description becomes a list of technologies. Culture becomes “we move fast.” Neither tells anyone who will fit. At this stage the fit that matters is founder-engineer fit, not company-engineer fit. It is how the two of you think, disagree and ship.
Mistaking a bar for fit.
“Top 1%” and “fewer than 3% accepted” measure how hard a test was to pass. They say nothing about whether an engineer takes ownership when nobody writes the spec, or whether you would enjoy two years of working with them.
I ran hiring the bar-first way for years, and it was right for the place I ran it. When I hired as a hiring manager inside a corporate environment at NBC, the process was built around two things: the technical bar, which tools like HackerRank measured, and fit with the organization’s culture. That works when there is an organization — the team, the conventions and the managers already exist, and a strong engineer who clears the bar slots into them. At a two-person startup none of that exists yet, so the bar tells you very little and the culture is whatever the founder does every day. A bar is someone else’s standard. Fit is yours.
Settling under pressure.
A recruiter sends three people and waits for a decision. The founder wants to see more, each shortlist round takes days, and the pressure to pick from what is on the table grows with every one. The fix is to slow down. Chesky, at Ramp’s Art of Hiring event in October 2024, told founders to “spend as much time referencing as you do hiring.”[9] A founder picking from a recruiter’s round does the opposite.
What founder-led hiring is supposed to be
Stripped down, founder-led hiring is three jobs that only the founder can do, and a pile of work that anybody can.
The three: judging fit, because nobody else knows what working with you is like; selling, because an engineer leaving a safe job needs to hear the mission from the person who holds it; and deciding, with the founder’s name on the outcome. Those deserve the hours Altman and Taggar describe, and they should not be handed to anyone. Not to a recruiter, and not to me.
The pile: finding people, finding out what they have built and how they work, scheduling, paperwork. None of it needs the founder’s judgment, and all of it eats the founder’s hours. This is the work a recruiting team absorbs later, and the work a founder does by hand now.
Done well, founder-led hiring is the founder spending their hiring time on the three jobs and almost none of it on the pile. Done badly, it is the reverse. Weeks in the inbox, ten minutes deciding. I have watched hundreds of founders do it the second way, and almost all of them thought they were being thorough.
When founder-led hiring stops, and what never does
Nobody has a published headcount for the end of the stage, but the most prominent founders I talk to agree on a rule of thumb. Lead every step of recruiting yourself for the first 50 to 100 hires, and spend at least 30 to 50% of your time on it from day one. The signal to hand over is a company near 100 hires and on track to grow 3x to 10x, which is when a talent director joins and reports to the founders. Others frame it by stage instead of headcount: lean on outside help only once the core team is built, and even then, a founder’s own pitch is the thing no specialist can replace.
After that, the work moves and the judgment stays. A recruiting team takes the pile of sourcing, scheduling, references and paperwork. The founder steps back from the early rounds and keeps the last one, the gate that decides who joins.
The founders who held that gate longest show how far it can go. Elon Musk personally interviewed nearly all of SpaceX’s first 1,000 workers, janitors and technicians included, and kept interviewing engineers as the company grew. Coinbase founders turned the same instinct into a process. Its leadership reviews every offer before it goes out, and its VP of Talent says knowing every hire reaches the founder changes which people his recruiters go after.[10]
Both stories make one point. Every later hire is judged against the people already there, so the early ones set the bar for everyone after them. A recruiter can match a brief. Only the founder knows what the company is becoming. My view is that the founder should keep the final interview long after a recruiter has taken over everything else.
Why I built TeamEx for it
I founded TeamEx in 2023 as a marketplace for hiring engineers, and for three years I watched the same thing happen across hundreds of founders and more than 2,000 engineer interviews. The founders who hired well kept the judging, the selling and the deciding, and had somebody else carry the pile. The founders who hired badly had either handed all of it away or were drowning in all of it. In October 2026 I rebuilt the company around that pattern.
TeamEx is tooling for founder-led hiring: search and interview scheduling, and the three jobs left with you.
TeamEx builds every engineer’s profile from evidence of their real work, and makes them available for founders to search and filter by 3,000+ skills and 400+ traits of how they work, think, and communicate. Think of it like a search engine, where you can search thousands of engineers using traits beyond what is on their resume or LinkedIn profile, each one tied to its source. You search on your terms, filter people by traits you would enjoy working with, interview, and decide on your terms.
Engineers build their profiles by connecting their work: coding-agent sessions, GitHub, Hugging Face, talks, writing and posts. They can add AI interviews and interviews with expert interviewers, each on one area, such as technical depth or how they work, think and communicate. TeamEx turns that evidence into facts with written rubrics, and every fact ties back to its source and date. Search ranks profiles by strength and recency. There is no pass or fail. That leaves the judgment, selling, and deciding to you.
Search for what you would judge in person.
No filter for it? Write your own.
The traits founders usually look for
Tap a trait to add it to the search below.
Founders often go from first search to the person they want to interview in a few minutes, 85% of hires are still in place at three months, and founders at Clipboard Health, Maple and TopHelpers have hired this way. That is the weapon founders at this stage need.

Zafar Ibrohimov
Founder, TeamEx
Notes
- [1] Harj Taggar, “How to Hire Your First Engineer,” Y Combinator (guide). The guide I send every founder before their first search. A third of your time is the floor, not the ceiling.
- [2] Paul Graham, “Founder Mode,” September 2024 (essay). Written about running a company. The hiring version of the same bad advice arrives two years earlier.
- [3] Chesky on his recruiting hours and Altman’s 50% rule, as reported by Fortune, September 28, 2026 (article). Three hours a day at Airbnb’s size. At two people it is more.
- [4] Sam Altman, “How to Hire,” 2013 (post). Short, and “you can’t outsource this” is the whole argument in four words.
- [5] Dover, LinkedIn Recruiter Lite vs Professional, 2026 (guide). The seat price is the giveaway: it assumes a person whose job is to sit in it.
- [6] Vendr, Greenhouse pricing (marketplace). A median contract, so half of buyers pay more. For two hires a year.
- [7] Paul Graham, “Do Things That Don’t Scale,” 2013 (essay). He was writing about recruiting users. Recruiting engineers is the same reluctance.
- [8] Nick Grandy on joining Airbnb as its first engineer, Y Combinator (post). A long search and still a take-home, two founder interviews and a beer. That is the right amount of rushing, which is none.
- [9] Chesky at Ramp’s Art of Hiring, October 2024, as reported by Fortune (article). Referencing is the part of the pile a founder should keep a hand in.
- [10] Greg Garrison, Coinbase VP of Talent, interviewed by Metaview (post). Every offer reviewed at the top, and it changes who the recruiters bother to chase.
