Why Great Hires Fail at Startups.

When Garry Hill recruited for SpaceX, interviews ran on Saturdays and Sundays. Push back on a weekend slot and you never got the interview. The calendar was the test.

People want to join startups for the same reasons people found them: ownership, speed, equity, the chance to matter. Most have never worked without the scaffolding of an established company. No recruiting team, no clear lane, no logo to borrow credibility from. That's why great hires fail at startups. Talent is rarely the problem. Their talent was just built for a different environment.

Garry has recruited at PlayStation, Amazon, Blizzard and Reddit, and is now the only person doing his job across Makers Fund's 130-plus portfolio companies. More than 20,000 interviews in, his view of who fits comes down to a different question for each side of the table.

Employees should ask whether they want the job or the idea of it. Employers should test what a candidate actually did, not where they did it. Founders should ask whether the person who fits today will still fit two stages from now, and that includes themselves.

Watch the full conversation: YouTube · Apple Podcasts · Spotify

Will a great employee from one company succeed in another?

Garry's answer is yes, but only downhill. Someone who made it at SpaceX would likely do fine at Amazon or PlayStation. The reverse rarely holds. SpaceX selected for mission obsession and tolerance for stress, so anyone who survived it had already been tested. For Garry, the through-line is grit.

"At Blizzard, the people who looked best on paper did the worst. Principal-level engineers and designers from top schools and leading tech companies struggled because it took a long time to build trust with the team. This top talent was typically used for quick wins. They often didn't have the grit to build trust. So they burnt out and spun out after about a year from Blizzard." — Garry Hill

His own Blizzard and Amazon stories from the podcast are about driven, pedigreed people who failed because the environment rejected them, and not because they lacked drive. Drive gets you in the door. Whether you stay depends on fit with how the company actually works. Fitting in takes grit and humility. So the useful interview makes sure the candidate understands the job they're walking into, and whether they actually did the work on their resume.

Regarding resumes, especially with big-tech hires, separate the work from the team. Ask which part of the work they owned, then back-channel the people who sat next to them.

Garry also named two types who underestimate startups. One is the visionary designer who forgets it's a business. The other is the strong engineer, allergic to anything businessy, who suddenly has to sort out the finance stack because nobody else will.

It's common to assume that startups are not for people with kids. Yet Garry pointed out that a family isn't automatically the dealbreaker people might assume. Parents can and do thrive at startups. What breaks the parent hire is dishonesty about the hours, from either the candidate or the founder.

Hire slower, wider, and with a process

The two mistakes Garry sees most in hiring are going both too fast and too narrow. Too fast means seats get filled because the founder, who's doing 8 things at the same time, wanted someone to fill the seat yesterday. Too narrow means hiring for titles rather than the role.

Referrals are the trap. Every startup fills its first thirty seats with "I know a guy." Garry doesn't say stop. He says build the process first: who interviews, what they ask, five steps maximum. Then run referrals through it like everyone else.

"A famous logo tells you nothing about how someone performs." — Garry Hill

The counterargument is real. Jordan Mazer, head of talent at a16z, pushes for speed because slow processes lose candidates, as he explained in our podcast. In California, with at-will employment, a fast mis-hire is cheap to undo. In most of Europe, it isn't. Your labour law matters as much as your philosophy.

In the early phase, you want T-shaped people, deep in one area and willing to cover the tangents. Because you're likely a pivot or two away from what your company should be building. Post Series A, when you have product-market fit, touching everything becomes the bottleneck, process arrives, and the CEO has to let go without pulling work back. Later stage needs people comfortable not being in every meeting, which is exactly where the scrappiest early hires struggle.

So what happens to the people who got you there? I think of them as rocket boosters. They do the hardest part of the flight, then detach, ideally thanked publicly. That's why Garry advocates for T-shaped talent. They are most likely to survive each evolution stage of the company.

Enjoying this? Subscribe to the Deconstructor of Fun newsletter for weekly deep dives on games, startups and the people building them.

Founders that turn into mascots or demigods

In scaled companies, lingering co-founders can act like demigods. No process applies to them, and they drag the company back to how things used to be. Garry agreed, then named the worse version: the co-founder who has no ownership and does nothing. A talking head. A mascot who supposedly represents "the culture." Left for everyone to wonder what they even do.

For co-founder candidates, Garry gets an hour and spends it probing for resilience. He asks for a story the person didn't choose: a lost job, a failed product. Do they own it cleanly, or blame the market, bad managers, no onboarding? Did they climb out, or did luck?

Then comes the move I'm stealing. Mid-interview, he asks who their mentor was. Then he asks them to spell the name. Or their lead on that feature, spelled out, because he's going to contact them. A reference in real time. Or an interview that falls apart because the candidate can't provide an answer.

"It becomes like a truth serum for the interview." — Garry Hill

On gaming versus tech founders, Garry nailed it. Tech founders solve measurable problems. Gaming founders bet on their own taste and go looking for fun, which nobody can measure up front. A thousand superfans is not a business. The founders who cross into tools and platforms pair grit with the humility to drop a thesis. For example, Heroic Labs founders fancied themselves as game makers 11 years ago. Turns out, they weren't good at it. Turns out gaming infrastructure was their game and it took humility to accept reality. Hear their founding story on our podcast.

A chip on the shoulder is fuel or poison depending on what it's made of. Garry cited Billy Corgan, singer of the Smashing Pumpkins, saying he wanted to out-earn the classmates who bullied him. What matters more to me is whether the founder can notice progress at all. When I interviewed Batu Çelebi of Grand Games, whose company reached a $500M run rate in about two years, he stressed celebrating early wins because nobody claps for you then. Garry once had a portfolio founder write every milestone down because nothing registered as success until it was on paper.

Finally, what separates someone who will build from someone who will quit? Garry says it's how long they've thought about the problem, and whether they're humble enough to change the path. Can those two people look identical in an interview? Of course. Interviewing is like dating. It's a less satisfying answer than I wanted, and I think it's correct. An hour won't separate conviction from performance. Real references, solid onboarding, and systematic performance review is the proven option.

Culture fit is the most common test done wrong

Every founder says they hire for culture fit. In practice, at startups, culture fit usually means two things. One is no jerks, which is fine. The other is hiring people from the same background and the same network. That second part is where bias creeps in, and teams start cloning themselves. And it works, until it doesn't.

"I like culture add, not culture fit." — Garry Hill

The way around self-cloning your team is asking what the team is missing and screening for that. Teams can't see their own blind spots, so Garry's fix is unglamorous. Survey the team on why they joined, what they love, and what's missing. The gap might be age or background. Or something as concrete as a studio full of PC players where nobody really plays mobile that much, all while working on a mobile game.

But what if you want to change your company culture fast? Say you want your team to work faster, and you hire a producer to speed up a slow team with ruthless aggression. Garry has seen it work. When a new hire is brought in as the hammer to move slow people out. It always fails. As a result, the people actually do less and leave.

A better way is to do cultural changes methodically. Provide clarity on where the culture is now, where it needs to be, and why. Then involve leaders in the transformation of the culture. Some will disagree and commit. Others will disagree and exit. And the culture will evolve.

My take: culture add is a smart way of thinking, but only if the founders can first describe the current culture in words. The problem is that most founders can't. Without that, the culture fit interviews are a waste of time. Vibe check cluttering CEOs' busy calendars.

The takeaways

Great hires fail at startups for predictable reasons. They were great inside a machine, and the startup has no machine. They were hired fast, through a friend, for a title, by a founder looking for someone familiar. And nobody told them the truth about the hours, the scope or how long trust takes to earn. If you're joining, test the job before you take it, fractionally if you can, and be honest about the life it requires. If you're hiring, run every candidate through the same process, ask what they owned rather than which logo they wore, and screen for what your team is missing. If you're a founder, accept that the people who got you here, possibly including you, may not be the ones who get you to the next stage, and plan for it before it becomes a mascot problem.

One of the first people Garry hired at Sony is still there, manager to SVP, one job for a whole career. After 20,000 interviews, that's the outlier worth remembering. Most hiring is managing mismatch. The best you can do is find it before it finds you.

Bonus: go fractional first

Garry takes weekly calls with senior people leaving big companies. He often pushes them toward fractional CFO and CMO work. Teams are runway-conscious, and a temporary senior hire is an easy yes next to a C-suite bet that sinks the company if it misses.

Having worked in a fractional role after exiting my own startup, it's the advice I give my senior friends looking for their next opportunity. If you're interested in getting an expert talent into your team for a fraction, or want to become a fractional talent, reach out!

Watch or listen: YouTube · Apple · Spotify

More from the podcast

Jordan Mazer (a16z) on the art of hiring and getting hired: YouTube · Apple · Spotify · Article

Joe Burridge (Play Ventures) with Miki Blasko on the hiring playbook of top-performing startups: YouTube · Apple · Spotify · Article