How to Hire Well: Bezos, Musk, Revolut, and Two Hires of My Own

Nothing you do as a founder or manager compounds harder than who you let in. A wrong hire costs you money, but a wrong hire in a key seat costs you time, culture, and the trust of everyone who has to work around them.

Why hiring is the highest-leverage thing you can do

The lone wolf doesn’t work. Not in tech, not anywhere, not ever. Every company you can name that reached real scale was built by a team, and the founder’s actual job, past the first few months, was hiring people better than themselves at the things that weren’t their strength.

Look at the supposed counterexamples. Steve Jobs needed Wozniak to build the thing, then needed Cook, Ive, and thousands of others to turn it into Apple. Bezos wrote the original Amazon code himself and then spent the rest of his career almost entirely on hiring, writing three-question interview frameworks and building a Bar Raiser program, because he understood scaling is a hiring problem. Musk personally interviewed SpaceX’s first 3,000 employees and cold-called top students in their dorm rooms, because a rocket company of one person produces zero rockets.

The “one-person billion-dollar company” line is a lie, one Sam Altman likes to float at conferences, a sales pitch for an AI product.

Bezos: raise the bar, every single time

Bezos’s hiring test, from Amazon’s early shareholder letters, comes down to three questions asked of every candidate:

1. Will you admire this person? “For myself, I’ve always tried hard to work only with people I admire… life is definitely too short to do otherwise.”
2. Will this person raise the average level of effectiveness of the group they’re joining? The goal is to fight entropy. Left alone, average quality regresses to the mean of whoever you hire next. Every hire should push the mean up, not just fill a seat.
3. Along what dimension might this person be a superstar? Not “are they well-rounded,” but “what’s the one thing they’re exceptional at that we don’t have yet.”

The mechanism Amazon built around this is the Bar Raiser: a trained interviewer, unrelated to the hiring team, with a specific veto over the hire if the bar isn’t met. It exists because “will you admire this person” degrades fast once a hiring manager is desperate to fill a seat. You need someone in the room whose only incentive is protecting the standard, not closing the req.

Musk: evidence over credentials

Musk’s core interview question, repeated at the 2017 World Government Summit and elsewhere, is deceptively simple: “Tell me about some of the most difficult problems you worked on and how you solved them.” The value isn’t the question, it’s what he does with the answer. People who actually solved the problem remember the ugly details: the dead end they hit, the constraint that forced the tradeoff, the specific number that told them they were wrong. People who were adjacent to the work, or padding a resume, stay vague at exactly the point where the real story gets specific.

His broader philosophy, from a 2014 interview: he looks for “evidence of exceptional ability.” Track record beats pedigree. A degree is a weak proxy for capability; a demonstrated, specific accomplishment is a strong one. This is also why Tesla and SpaceX have both hired people without degrees when the portfolio or track record spoke for itself: the credential was never the point, it was always a proxy standing in for evidence that happened to be easy to check on a resume.

But talent alone isn’t his whole answer, and to his credit he’s said so publicly. Asked at SXSW about the biggest mistake of his life, Musk didn’t point to a technical failure. He pointed to hiring: “The biggest mistake, in general, I’ve made, is to put too much of a weighting on someone’s talent and not enough on their personality. I think it actually matters whether somebody has a good heart, it really does.” He put it in team terms: the best person isn’t necessarily the one who scores the most goals, it can be the one who assists the most goals. Evidence of ability gets someone in the room. Character is what determines whether the team is still functional a year later.

Revolut: the machine, not the magic

Revolut’s own playbook (built with QuantumLight) goes further than a hiring philosophy: it’s an operating system for hiring at volume without diluting the bar. The short version:

• Exceptional talent beats large teams. Density scales, bureaucracy doesn’t.
• Hire ambitious juniors over “experts.” 7-8 years of experience for leadership, 2-3 for individual contributors: people early enough in their trajectory to still be climbing it. They report replacing several tenured executives with junior, hungrier people who were simply more willing to get their hands dirty on problems they’d never solved before.
• Build an internal recruiting machine, not a dependency on external recruiters, because external recruiters optimize for filling the req, not for your five-year talent bar.
• Hyper-structure everything. An unstandardized process decays under volume. Consistency is what defeats bias at scale.

Their concrete mechanism is a three-interview framework used for nearly every role:

1. Problem-solving assessment: a multi-layered case study with 2-3 root causes, given with zero data up front. The candidate has to ask for the data they need, which tests quantitative reasoning as much as the final answer.
2. Bar raiser: digs into career progression, failures, and metrics, looking for actual proof of being a recognized top performer, not a polished narrative.
3. People management assessment: for anyone hiring others, a case study specifically on how they’d protect the talent bar as a manager, since a hiring manager who won’t hold the line undoes everything above.

And their 5-step bootstrap sequence if you’re starting from zero: hire your first internal recruiters, appoint role owners who define what “good” looks like per function, define the company-wide bar with the two non-negotiable interviews above, write standardized playbooks so quality doesn’t depend on who’s interviewing, then embed the whole thing into your ATS so pipeline health is a metric, not a vibe.

Masterclass hiring moments, ranked

A few of the best hiring moves ever pulled, roughly from most to least impressive. Some of these are cold and calculating, some are pure conviction. Worth noticing which is which.

1. Chase the whole cohort, not just the ones who found you. A professor once surveyed his ten best students to see what made them exceptional and published it. Six of the ten had gone straight to SpaceX. Musk didn’t send a thank-you note, he arranged to meet the professor and, once there, asked for the names of the other four. He wanted the full set, not just the ones who’d already self-selected in.
2. Make first contact yourself, before you have any credibility. In SpaceX’s earliest, least credible days, Musk personally called top aerospace and physics students based on nothing but grades, sometimes reaching them in their dorm rooms. One recruit assumed it was a prank call and didn’t believe the company was real until he looked it up. Credibility is usually the hardest part of recruiting, and sheer personal effort can substitute for it.
3. Remove the real blocker yourself instead of arguing around it. A well-known version of this story has Musk wanting to hire someone who wouldn’t relocate because their spouse had a strong job tied to a specific city, so he picked up the phone and personally worked the spouse’s situation until it stopped being a blocker. I couldn’t pin this one to a documented primary source, so treat it as widely told rather than confirmed, but the mechanism holds regardless: don’t let a solvable logistics problem cost you the hire, solve the actual problem.
4. Sell the mission in one sentence, not the business plan. After Apple’s board spent two days failing to close a Fortune 500 CEO with numbers and slides, Jobs took one meeting himself and asked one question: “Do you want to sell sugar water for the rest of your life, or do you want to come with me and change the world?” One sentence did what two days of business plans couldn’t.
5. Watch how they think, not just what they answer. Amazon executive Ann Hiatt describes being hired on the spot after Bezos gave her a live brainteaser (estimate the number of panes of glass in Seattle) and watched her reason through it out loud. He wasn’t grading the final number, he was watching the process.
6. Make poaching irrelevant by paying before it happens. Reed Hastings’s Netflix told employees to take recruiters’ calls and report back the offers, so Netflix could out-pay them before they left, reasoning that hiding from the market only delays the moment your best people walk anyway.
7. Bet on character with zero traditional vetting. Nike’s Bill Bowerman almost never got involved in hiring, but one day called Phil Knight and simply told him to hire Bob Woodell, a promising college runner who’d just been paralyzed in an accident. No process, no interview loop, just a character judgment from someone who knew him. Woodell became one of Nike’s most important early operators.
8. Total conviction can outweigh a thin resume. Jeff Johnson, Nike’s first full-time employee, ran the entire early operation out of the trunk of his car: selling shoes at track meets, writing customers personal letters, designing shoe models in his spare time. He’s also the one who came up with the name “Nike,” reportedly in a dream. The hire wasn’t a resume match, it was total conviction meeting total trust.

Two hires of mine, same instinct, opposite outcome

Here’s where I’ll admit theory and practice don’t always line up neatly.

The good one. I once hired a friend’s friend, vetted mostly through one conversation and the fact that someone I trusted vouched for him without hesitation. Easiest hiring decision I’ve made. Immediate trust, no ramp-up drama, no surprises. A strong referral is a shortcut to signal, not a shortcut around it.

The bad one. I hired someone purely on gut feeling and good intent, and deliberately skipped any competence or reference check because it felt generous and human at the time. He didn’t get his work done, was consistently late on commitments, and in our syncs he would always warn me about incoming “UFO attacks.” Not a joke, he meant it 100% seriously. It was an unpaid, volunteer role, so the damage was contained. Zero verification doesn’t just risk a mediocre hire, it removes your only early warning system.

The lesson isn’t “don’t trust your gut,” since the good hire was gut instinct too. It’s that trust needs to be trust in something specific: a referral you can weigh, a track record, anything you can actually check. Trust with zero evidence behind it isn’t a signal, it’s a coin flip, and eventually it comes up UFOs.

Putting it together

Across Bezos, Musk, and Revolut, the same shape keeps showing up: define the bar explicitly, test for real evidence instead of proxies, and put a structural check in place that survives the moment a hiring manager gets desperate to fill a seat. My own two data points say the same thing from the ground level: the hire that worked had real signal underneath the good feeling, and the hire that didn’t work had none.