Khosla Ventures

People decide the outcome more than market, product or technology. Vinod Khosla gave Keith Rabois this line while sitting on Square's board. Rabois had already seen it at PayPal: with the right people everything else is easy, and with the wrong people everything is hard.

Founders get distracted by market, product and technology. Rabois adds that a founder who can judge talent ruthlessly and accurately can go very far with no other abilities whatsoever.

Lenny's Podcast

PayPal's dense talent pool built PayPal and seeded a generation of companies. Max Levchin hired nearly all the technical talent and Peter Thiel hired almost everyone else, mostly through first or second degree links to the engineers or to Thiel through Stanford. That talent went on to found companies in many verticals over 25 years.

Rabois would not recommend network hiring for everybody. It works only if your network holds unique talent, and then he sees no substitute.

Lenny's Podcast

Output is capped by the number of barrels, not headcount. A barrel is someone who can drive an initiative from inception to success. PayPal had about 254 people in Mountain View and roughly 12 to 17 barrels; Lotus had two.

"If you hire more people without expanding the number of what I call barrels that can drive from inception to success, all you're doing is stacking people behind the same initiatives." The counts are asserted from experience, with no data shown.

Lenny's Podcast

Post-raise hiring sprees disappoint when only ammunition gets added. After a Series A or B, companies hire fast, burn more and get frustrated that little more ships, because extra hires behind the same initiatives raise the coordination tax. The number of barrels sets how many things can run in parallel or in sequence.

Ammunition still matters in a proportion that fits the project: some need a designer, engineers, a PM and a data analyst, while a single barrel can succeed on others with little or none. Thinking in ratios makes team construction deliberate.

Lenny's Podcast

A trivial task can reveal a barrel, as Square's smoothie test did. Rabois wanted cold, healthy smoothies delivered to late-working Square engineers at 9 p.m., and his large office team and EAs kept failing. Intern Taylor Francis, on his second day, said he would solve it, and the smoothies arrived cold at 9:00.

A barrel motivates people, gathers resources and measures progress, and can be fired and forgotten; if it stalls, they come back early with the issues, what they tried and a root-cause diagnosis. Rabois's reaction: "I found a barrel."

Lenny's Podcast

Startups win by finding undiscovered, underpriced talent. Large recruiters process people through a homogeneous black box, and younger people have fewer data points, so they get misjudged and underpriced. Rabois got promoted at PayPal by recruiting under-used people already in the building after his outside hires came out about 50/50.

COO David Sacks told him he would not advance until he showed leverage, where 1 + 1 equals three or more. His read worked on people he knew from lunches and runs but failed in a 20 to 45 minute interview, so he spent years building a method for strangers.

Lenny's Podcast

Reference heavily and keep going until you hit a negative. David Sze of Greylock taught referencing a founder until you get a negative reference, and DoorDash's Tony does 20 references per senior hire. Ask whether someone can be a world-class entrepreneur, not whether they were a good employee.

Rabois probes two arcs: what makes the person most successful, and the root cause if they fail. References fit poorly for fresh college graduates.

Lenny's Podcast

Ask senior candidates what they'd do differently as their CEO. The question shows how much of the business model they absorbed and whether they grasp tradeoffs. He follows with: why couldn't you persuade the CEO to do it?

It is not meant to bash the employer.

Lenny's Podcast

After 30 days, ask the whole team whether they'd hire again. Rabois cites research that this 30-day answer is about as accurate as one taken a year or two later, and says every company should use it. Ask the whole hiring team, not only the manager.

Hiring is a muscle: try, review what you missed and why, repeat. Tactics move you 10 to 20 degrees within the bell curve, but finding the top 10 to 50 basis points of talent means deviating from the norm.

Lenny's Podcast

Top talent joins when their skill removes your biggest bottleneck. The best people come for a big mission and when their own skill is the thing the company is missing.

Lenny's Podcast

Back founders with a non-zero chance of changing an industry. Rabois's early-stage test is whether the founder could change an industry, and he wants half his VC friends to laugh at a seed deal. His early investments include Stripe, Palantir, Airbnb, YouTube, DoorDash and Ramp.

Lenny's Podcast

Ask AI founders where accumulating advantages will come from. Check whether foundation labs will leave room and whether the business gets better and easier over time; network effects are only one type. At seed, the founder only needs to say where advantages could build and when they'd start measuring them.

"You want something that over time makes the business better and better."

Lenny's Podcast▶ 1:04:20

In AI, product judgment beats process and consumer feedback misleads. Year-long roadmaps are incoherent when models change monthly, so the scarce skill is deciding what to build and why, like a CEO. Consumers explain subconscious choices with false reasons; Airbnb would have failed a survey of ten random people.

The advice comes from a survivor-heavy view of PayPal-network and venture-backed winners and does not weigh the cost of dismissing customer feedback.

Lenny's Podcast

Keep tempo high and push harder as results improve. Complacency grows with success, and top performers lose morale when the team coasts. Ramp was close to shipping cards in 3 months, against a normal 9 to 12.

Lenny's Podcast