Keith Rabois (@rabois) is an investment partner at Khosla Ventures. Since 2000, he has been instrumental in driving five startups from their early stages to successful IPOs, with executive roles at PayPal, LinkedIn, and Square, and as a board member with Yelp and Xoom.
At KV, Rabois has led investments in a broad array of startups including Door- Dash, Stripe, Thoughtspot, Affirm, Even Financial, and Piazza. While working as a VC he simultaneously cofounded Opendoor, a startup in the real estate tech world.
I sat down with Keith Rabois to talk through some of the nuts and bolts of pushing a hypergrowth company to the next level: When (and why) to IPO, how to find stellar executives, and why a lot of founders need to pare down their roster of direct reports.
Elad Gil: I think a lot of founders struggle with how to hire their first CFO, their first general counsel, even their first even VP engineering. Those may be functions they’re unfamiliar with, or functions where they don’t really have a network. I guess the first part of hiring is knowing what “great” is for a role or function. How can a founder who’s never done it actually know what’s great?
That’s a very significant challenge for people hiring outside disciplines that they’ve actually worked in. So a great designer tends to know how to hire a design lead. A great engineer may know how to hire a VP of engineering. But they may not even know what a CFO does, let alone what a great CFO is versus a good one versus an excellent one.
One technique I learned, actually from Brian Chesky at Airbnb, is to go find the five best people in Silicon Valley that do that role, and just have coffee with them. And just chat. In that dialogue, I think you form an ability to benchmark the differences between an A+ and a B+, so that when you meet new candidates, actual candidates, you can triangulate against the people you’ve met that are clearly stellar. And so you should use your board, your investors, any connections you have to get introduced to the five best people, and then spend some time.
Secondly, if you have great investors or board members, absolutely they probably have some experience in hiring some of these roles. And you want to rope them into the interview process earlier rather than later. That can be a high-leverage move.
But it is challenging. You may be able to find a friend or colleague or comparable founder who actually grew up with a different background than you that you can trust, and ask him or her to interview people. Back in the day, after I left PayPal and was doing some other startups, the biggest challenge for me, not being technical, was to hire a VPE. So when I’d get to finalists for VPE positions, I’d ask Max Levchin, who’s an extremely strong technologist, to interview the one or two finalists and give me feedback. Obviously I couldn’t borrow all of Max’s time, and I couldn’t constantly ask him to interview people. But when I’d get down to one or two choices and it was a really important hire, I’d ask him to send me feedback.
Elad: Outside of either board members or venture investors, how else would you go about sourcing an executive? Would you use a recruiting firm?
Keith: I would. For senior executives, executive recruiters can be very beneficial. And what I mean by that is if it’s a C-level officer, or a VP or above, executive recruiters really know how to surface candidates. They don’t know how to get you directors or more junior managers. But if you’re actually aiming at the high end, I think an executive recruiter can be very beneficial for two reasons: One, they have a network. They know who’s on the market, who’s looking around for new opportunities. They also know reputations. They’ve probably performed reference checks on these people before. And you can avail yourself of that.
Two, they will install some process. Just the discipline of having weekly meetings can accelerate your search. So I think that’s a very good idea. I highly recommend it for senior hires. It’s not that expensive in the grand scheme of things. You’re talking about $100k, which, if you’re hiring a great CFO or VP of engineering is totally worth it.
A lot of the VC firms that you may be working with also have internal executive recruiting and talent partners that are high-end executive recruiters, or were. So you can leverage them for free. For example, we meet with senior executives at companies as they’re starting to think about moving on. And we will figure out what their criteria and skills are and try to repackage them and recycle them into portfolio companies. So that can be a free way of jump-starting the process, which is not a bad place to start.
In addition, these days, use social media like Twitter. You know, if your company is doing really well, post that you’re hiring a CFO. You never know who responds that you might have been afraid to approach. I see some of our companies having great success with tweets about “I’ve got this great opportunity.” So there’s a lot of room for creativity as well. But I think the basic blocking and tackling of using an executive recruiter can work.
Now, there are some disadvantages of using an executive recruiter that are worth highlighting. Their incentive is to close a deal; they get paid when you hire somebody. So they’re going to want for you to hire somebody. And they may, if they’re not an awesome executive recruiter, be biased toward people who are easier to close. Just classic incentive alignment. They will find people that are easier to close than your dream candidate. Your job as a CEO is to get your dream candidate, and it may take you a year to get your dream candidate sometimes. So there is a little bit of a misalignment there that’s worth tracking.
However you source candidates, one key criteria for any executive is their ability to attract and become a magnet for talent. So when you’re doing reference checks, you really want to understand: Does this person have a pool of amazing people that will want to join a company? Can they immediately upgrade the entire talent of the organization because they’re so talented that other smart people really want to work with them?
That’s something you can tell from reference checks. And you should thoroughly do reference checks on executives, without fail. There are sometimes excuses why individual contributors don’t get reference-checked to death. But with executives, there’s no reason ever for hiring someone without thorough reference checks.
“One of the things you should always ask is, ‘If this person joined my company, would you join?”
– Keith Rabois
So, when you talk to various colleagues of this person, one of the things you should always ask is, “If this person joined my company, would you join?” And yeah, people may say, “I’m retired” or “I’m a VC now,” and all these reasons. But fundamentally you should hear in their voice, “Absolutely, yes.” At least some of the time. It doesn’t have to be a hundred percent of the time, but you want to hear that. And if you don’t hear that spark, there’s probably something there that’s worth really probing into.
Elad: One of the things that I tell founders often is that they should give themselves permission to screw up when making an executive hire, because I think fear really stymies or prevents people from going ahead and doing it. What do you think are signs that somebody is working or not working out as an executive, and how quickly can you actually tell?
Keith: Usually for an executive, it’s pretty reasonable to know in 30 days, certainly by 60. It does depend upon the complexity of the business. Some businesses are incredibly complex, like, let’s say, Opendoor, where even the best executives in the world take a little bit of time to really master how everything connects. But typically executives are super savvy. The pattern recognition that they’ve developed over their careers enables them to cut to the chase quite quickly. And if you see them struggling early, it’s often a major red flag.
Now, that said, your job as CEO or founder is to help them be successful. I think taking the obligation personally and doing everything in your power to make your new executive successful is part of your job. And some founders don’t do that. They just assume, “I’ve hired this person and this person will get up to speed and start doing things.” But I think carving out 10, 15, maybe 20 percent of your calendar to help make this new executive successful is an incredible investment actually, with very high dividends.
Because there is some expense. If you’re wrong, there is definitely some pain and friction in replacing the person. That said, it’s rarely fatal. Lots of companies, lots of founders, have made mistakes at the executive level, and upgraded. So for example, Mark Zuckerberg basically replaced his entire management team starting in 2007. By 2007, Facebook was a very successful platform and company. But there is literally nobody left on the executive team that was an executive in 2007.
So clearly you can constantly look to improve. You don’t have to find the magical solution right away. If you aim for zero-defect hiring, it’s a little bit like zero-defect decision-making: you’re probably too conservative. We teach, and I subscribe to the view, that you want to pull the trigger on an initiative or an executive hire when you’re about 70 percent confident that it’s the right decision. Below 50 percent is kind of reckless. But if you go for 100 percent, you’re waiting too long and you’re probably losing candidates. Your false positive rate may be low, but your false negative rate is going to be very bad, too. And a lot of people hiring don’t track their false negative rate of people they didn’t get that they should have gotten.
Elad: Are there any early signs of somebody not working out?
Keith: Yeah, usually the signs are that they don’t take ownership of decisions. Now that can be somewhat CEO- and founder-driven, too. Sometimes giving them license to start doing stuff requires a direct conversation. They may be a little bit too nervous of rocking the boat. So that’s one, they’re too passive, in effect.
The other is when people start circumnavigating around them and back to you with problems. That can be a sign of two things: It can be a sign that the executive isn’t doing well. It can also be a sign that you made a political transformation, which isn’t the easiest to disentangle. Obviously if you made a significant leadership change, there will be people who want to work within the old regime. And you may see some of that show up on your desk, so to speak. But often it’s also a sign that the executive isn’t doing well when people are still coming to you to problem-solve in areas that the executive should be managing.
Ultimately, the way I grade people is pretty subtle. You can usually look around an office, especially if it’s an open office, and see who’s coming to people’s desks. The people who are thriving—at any level, junior to senior— tend to have people approaching their desk all the time. Because it basically means that someone is going out of their way because they believe that this person can be helpful. And so insofar as people start going to this executive frequently, even people beyond the organization that the person is managing start approaching them and working with them and meeting with them, those are really positive signals.
“The people who are thriving – at any level, junior to senior – tend to have people approaching their desk all the time.”
– Keith Rabois
Another lesson that I learned from Brian Chesky—one way to think about when to upgrade executives—is that a really great executive is about six to twelve months ahead of the curve. They’re already planning for and acting on things that are going to be important six to twelve months in the future. A decent executive is delivering in real time, now to one to three months in advance.
So you can start measuring your executives that way. Are they seeing around the corners of whatever they need to? Because not everything can be changed overnight. So let’s say you needed fifty more engineers. You can’t hire fifty engineers tomorrow. But a great VP of engineering or someone more senior may realize that the strategy requires us to have fifty more engineers, or there’s no way we’re going to be able to deliver on what we need to deliver. So they start recruiting a year in advance. Things like that.
Elad: They’re really thinking far ahead on the curve. I guess the flip side of that, though, is the exec who’s five years ahead of where the company should be.
Keith: Yeah, that doesn’t work. That’s why I think the six to twelve months is about right. One to three months, you can leave them in place but they’re probably not ideal. Six to twelve months is a superb executive.
And it may look flawless. You actually have to know what’s going on very precisely. To many people in the organization, the executive may just look flawless, because they always have an answer. That’s partially because they’re six to twelve months ahead in their own brain. They know what’s going to break. They know what’s going to take time to fix. And so by the time it starts showing up in the organization, they can say, “Aha, we’ll just do this, this, and this.”
It’s like great software architects for companies. When they have scaling issues, they already have their silver bullets in their pocket. They’ve already thought through, “If we went on Oprah today and got 10X the amount of traffic concurrently, what would we do? I’d do this first. I’d throw this CDN at it, then I would do this. And I’d have this hardware over here. Servers take long lag time, so I’d have to have those.” They have all that stuff in the back of their brain.
It’s just like that for business problems. Under the worst possible circumstances or the best possible circumstances that I foresee in the next three to twelve months, what are the lead times associated with those various things? So I have those tools at my disposal when I need them, and I can just snap my fingers.
Elad: How many mistakes can you tolerate in executive hiring? Is it a certain number per role? Is it a certain number of executive mistakes per year?
Keith: I think it’s about one. You’re only going to add so many executives. Most companies are probably adding one to three in a compressed period of time. If you start having multiple mistakes, I think there may be something wrong in your process. One is perfectly normal and natural. But if you start seeing multiple, then I would think through, “What am I doing wrong?”
Elad: And by “multiple,” do you mean over a certain period of time? Because every 12–18 months, many companies have to upgrade their team if they’re growing very fast. And some of those execs will continue to go with that growth, but some will also break over time.
Keith: Let’s say, in an eighteen-month or two-year time frame, one clear mistake. There’s a difference between a mistake and someone who doesn’t scale. I would separate those into two camps. So the mistake is, “Oh my god, I just need to make a change. This wasn’t the right decision.” That’s something you only want to do about once.
The person who’s not scaling to the new level of complexity of the business and the size of the team and the different kinds of problems, that’s different— that’s not necessarily a mistake. You would possibly have made the same decision if you had the same information you have now. Whereas the mistake, if you had the same information, you would not go back and make the same decision. So that one, you definitely want to limit to one digit if you can.
The scale and upgrade is also a function of the growth rate of the company. So the faster the company is growing, the velocity of that change has a different slope. And the executive’s learning curve may be below that slope. Then you might have to make a change.
Elad: Now let’s say that you’ve built out your executive team as a CEO. There’s a huge transition when going from 20 or 50 people to hundreds as CEO, in terms of how you actually manage your own team. How do you think a CEO should think of their team overall? Who should report to the CEO, and who should not report to the CEO?
Keith: It’s a great question, and I don’t think there’s a paradigmatic answer that applies to all companies. It depends upon what the company does and where what I call the seams in the business are.
Depending upon what types of decisions have trade-offs, you want those— the trade-off decisions—to be made very infrequently by the CEO. You don’t want to have to be a tiebreaker every day, or every week even. Once a month or once a quarter, it’s fine to tiebreak between different parts of the organization. But anything more frequent than that, you need to unify the functions underneath somebody. So that’s one thing I would look for.
Second thing is just skills. Different executives come with different strengths and weaknesses. And sometimes, even if on paper you draw a perfect org chart, an executive doesn’t have a particular skill but is awesome at other things. So you might make some compromises in your org design to reflect their strengths.
For example, I have a friend who’s a leading product guy who happens to understand biz dev and partnerships really well. That’s actually an unusual combination. So if he was a senior executive, I would throw partnerships and complicated negotiations with music studios or something like that under product, even though that would not be the standard design. He’s just extraordinarily good at it. So there are times when you can do that.
It does require a diagnosis of your organization. Also, what are the key risks to the business? What are the most important two or three things? You want the things that are one, two, or three to report to the CEO. Fundamentally, what makes or breaks the company should probably be reporting to the CEO. Because ultimately the CEO is responsible and accountable for everything. So if you do the rank prioritization of what gets you from point A to super successful, what are the two or three key levers? You want those two or three levers pretty close to your span of control.
Elad: What do you think is the right number of reports? Obviously it’s contextual and it varies by the skillset and all the rest. I’m just curious roughly what you think is a ballpark.
Keith: The traditional advice derives from High Output Management—you know, Andy Grove in 1982—which is at most seven. Or five. Five or seven. The reason why is that you want to do a certain number of 1:1s at a weekly pace, and you want to do about one a day. So that means five or seven or something like that. I think if you can get down to three to five, that’s ideal. If you have to go four to seven, that’s definitely possible to scale. So anywhere in those ranges can work.
Elad: I feel like a lot of the executive teams I see these days have a dozen people reporting into the CEO.
Keith: That’s crazy. That’s totally crazy. That being said, you may have to deal with that for a period of time as you’re recruiting someone to unify some functions.
For example, there was a point in time at Square where I had somewhere between 11 and 13. And, you know, Jack was furious with me about it. My board was furious with me about it. And I was totally aware that it was not a sustainable situation. But rather than artificially unify things, I wanted to find people that would start bringing together some of those direct reports. And I eventually did. But I figured rather than make a mistake and have someone report to someone where they wouldn’t be able to create value, I would deal with it on a temporary basis. But I would urgently be recruiting.
So, I think you can have periods of time—measured in months, not years— where you violate the rules. But that becomes your number-one priority: to stop violating those rules.
Elad: I think the biggest fear that founders have when consolidating their teams is flight risk. They think that if they layer somebody or somebody else comes in, employees will leave. What do you view as the best ways to mitigate that? Do you view that as a real issue, or is it okay if that person ends up leaving?
Keith: It is a definite concern, and I think it is the biggest reason a lot of founders procrastinate. The standard I was taught, which I subscribe to, is you only want to layer someone if the person above them is clearly superior.
One of the techniques if you have too many direct reports is to move somebody to report to another person that’s already in the organization. The problem is, unless there’s a clear separation between them, it just doesn’t work. It’s not fair, and it will create flight risk. But if there is a clear separation, where the performance is differentiated enough—looking at both the way you measure it and how other people perceive the two executives—then I think that can be a smart move. And then you don’t have the organ-transplant issue of bringing in a new executive. So that’s possible, but there has to be a significant delta in performance. If it’s even a debatable delta in performance, it’s not a good idea to ask somebody to report to someone else.
If you’re going to hire someone externally, I think generally the way to retain people who are performing and who you really want to retain is to hire someone that they can learn from. And if that’s true, a lot of well-motivated people will stay. Where they don’t perceive that they can learn from the new executive, it may be better for them, psychologically or professionally, to go somewhere else and get on a steep learning curve again. So it does depend. But I don’t think it’s a reason to avoid hiring.
“The way to retain people who are performing and who you really want to retain is to hire someone that they can learn from.”
– Keith Rabois
At the end of the day, if one of those 11 or 12 executives is so valuable to you, you can do a couple things. One is to put them on a mentoring path and try to teach them how to accelerate their growth so that they can handle the larger function. But it depends on the company’s velocity whether that’s possible. It also requires you to have a network of mentors that you can throw at this problem, and you can’t do it with multiple executives at the same time. So if I had a high-potential VP of finance and I really wanted to groom him to be CFO, it’s possible to do that rather than layering him behind the CFO. But it takes a lot of energy. I can’t do that with my director of product, my director of engineering, and my VP of finance at the same time.
Elad: How should the CEO run their own team meeting?
Keith: That’s a great question, because I think everybody gets frustrated by this topic at some point. One of the lessons I’ve learned is the executive team meeting is not necessarily just for the CEO. It’s actually often more important and more valuable to the executives that participate in the meeting. They get to be on a stage with their peers. They get to understand what’s going on laterally in the organization so they can make smarter and better decisions.
The meeting may not be particularly insightful to you, because you’re doing 1:1s with all these executives and functional organizations. You may know everything that comes up in this meeting, and you probably should. But it’s the debate or dialogue or the lateral sharing of information that makes that meeting constructive.
I think a lot of CEOs get frustrated because the meeting isn’t adding value to them. But they forget that it’s adding value to the three or four or five other people there. And if you can make your executives more successful with an hour of your time or two hours of your time, that’s totally worth doing. It’s a classic high-leverage activity.
Elad: What do you think is a great agenda?
Keith: I think limiting it to about three topics. There’s information sharing, but then there’s discussion and debate topics that are actionable. And I think people’s attention span sort of wanes if you have too many debates at the same time.
Some organizations allocate a full day to this, and they go very deep. I’m not sure that’s a great idea. It does depend upon how tightly aligned your organization needs to be to function well. I tend to think one to three hours is more than sufficient, especially if you have tools in place—metrics tools, KPIs, dashboards—so that everybody really understands the business before they’ve even shown up to the meeting.
In that case, I tend to prefer circulating notes the night before. They can be bullet point-style, using the three Ps—plans, progress, problems—and shared in advance, so that people’s brains are chewing on what’s going on. And then there’s a couple discussion topics that affect the organization, or where the CEO wants broader input because he or she doesn’t actually know what the right answer is.
Elad: The other thing I’ve noticed is that early on in the life of a company, a lot of topics exist in the one executive team meeting. And then later they start to get split out. There’s a rev force and metrics meeting that runs separately, which includes a subset of the executives. Plus there’s a broader team to talk about progress in the business, which may differ from some of those strategic topics in the executive team meeting.
Keith: Yeah, sometimes you separate out strategic topics that are going to be controversial, or topics that don’t necessarily have right answers but are a question of trade-offs. That’s different from an operating review, which is: What are our current KPIs? How well are we doing? What’s the rate of progress? What are the different things we should be doing? Sometimes you’ll have the first meeting run by the CEO and the second meeting run by the COO or equivalent officer.
This interview has been edited and condensed for clarity.