High Growth Handbook: Scaling Startups From 10 to 10,000 People

by Elad Gil

If the high-growth stage at your company feels like a chaotic, scary, stressful shitshow, don’t worry. It feels that way for everyone the first time around. Buckle up and enjoy the ride!

One of the things you see crystal clearly in VC is how much competition emerges whenever anything works. Every single time we say, “Oh, this startup is unique. There’s some unique product here and there’s not going to be competition,” invariably six months later there are 20 venture-backed competitors doing the exact same thing. And so at some point, if the early guys don’t get to the other 95% of the market, somebody else is going to go take it away. And whoever has 95% of the market, number one they’re going to get all the value. All the investment returns, all the employee compensation flows to that company. And then number two, that company then accretes resources so they can work backward. In a lot of cases, they end up buying the company that got the early adopters for a small percentage of their equity, and then they just take the whole thing.

companies that charge more can better fund both their distribution efforts and their ongoing R&D efforts. Charging more is a key lever to be able to grow. And the companies that charge more therefore tend to grow faster.

That’s counterintuitive to a lot of engineers. A lot of engineers think there’s a one-dimensional relationship between price and value. They have this mental model of commerce like they’re selling rice or something. It’s like, “My product is magical and nobody can replicate it, and I need to price it like it’s a commodity.” No, you don’t. In fact, quite the opposite. If you price it high, then you can fund a much more expensive sales and marketing effort, which means you’re much more likely to win the market, which means you’re much more likely to be able afford to do all the R&D and acquisitions you’re going to want to do. And so we always try to snap people into a two-dimensional mindset, where higher prices equals faster growth.

What I’ve always found is this: give me a great product picker and a great architect, and I’ll give you a great product. But if I don’t have a great product manager, a great product originator—it used to be called a product picker—and I don’t have a great architect, I’m not going to get a great product.

I would take more of a micro-view of it. Which is: Okay, how many great product pickers do you have, people who can actually conceptualize new products? And then how many great architects do you have, who can actually build it? Sometimes, by the way, those are the same person. Sometimes it’s a solo act. And sometimes that’s the founder. As you scale, you need more of those people. But I always think it’s a matter of, okay, how many of those people do you have or can you go get? Or, back to acquisition, how many of those can you acquire? And then basically that’s the number of products you can be working on. You organize R&D around that, in my view. You want to have a relatively flat R&D structure. You basically want to have autonomous teams, where each team is guaranteed to have a great product person and a great architect. And that’s the model.

I generally think matrixed is death, so I’m always pushing companies to go to a flat structure of independent teams. I’m really on the Jeff Bezos program on that, the two-pizza team thing.2 I think hierarchies kill innovation for the most part. And I think that matrixes are just lethal in most cases. There are exceptions, but in most cases, you need original thinking and speed of execution, and it’s really hard to get that in anything other than a small-team format, in my view.

Acts as chief psychologist of the company. Founders are often surprised by the extent to which people and organizational issues start to dominate their time.

If your involvement is not uniquely crucial to the success of a task, or an item is not core to your personal life, you should figure out how to off-load it. In most cases, that will boil down to simply learning how to say no,

First-round interviews. You don’t need to be part of every first-round interview for every candidate. You can still talk to people in later rounds or as a final sell. Executive hiring is different and as CEO you may need to actively reach out to candidates.

One key is to hold skip-level meetings without your reports feeling threatened. If a person’s manager shows up to what is explicitly a skip-level meeting, there is likely something wrong with that manager or their management style. It is also up to you to clearly articulate to your report that you hold skip-level meetings with numerous people at the company so that they do not feel threatened or concerned.

“Working with Claire.”

I think that founders should write a guide to working with them. It would be one of the pieces I’m describing, to clarify the founder’s role: “What do I want to be involved in? When do I want to hear from you? What are my preferred communication modes? What makes me impatient? Don’t surprise me with X.” That’s super powerful. Because the problem is, people learn it in the moment, and by then it’s too late.

would say that it’s a combination of what I would call “operating structures,” which are things like documenting your operating principles and processes. Operating structures are not tied to any one particular process, but instead explain, “This is what we expect of ourselves, in terms of how we work.” And when you document things like that, new leaders, new managers, new people in the organization can say, “Ah, this is what’s important. Let me adapt my behavior as we scale to follow along in this structure.”

One thing I’ve observed is that you can’t make too many things at a company mandatory. You really have to be judicious about the things that you’re going to require, because there just can’t be that many. There’s probably something related to performance and feedback. There’s probably something related to whatever your planning process is. And then there’s a few day-to-day tactical things, like a launch review. But you can’t have that many and you can only have one at each level.

As you scale, you realize, “Huh, I really need more of these.” And the danger is getting too process-y instead of outlining the objectives so people understand, “Okay, we’re doing this for this reason.” It’s almost like you want to provide more context versus trying to exert more control.

They need some structural boundaries, but you can’t over-constrain them. And that’s why you want to have some high-level metrics that everyone’s steering toward, operating principles, a documentation of plans, and then a set of processes that you follow.

Right, and just saying, “This is about context and communication, not about control” is so important. You literally have to state the obvious and make sure people know that.

Sometimes that gets confused, though, because the group thinks they’re the decision-maker and that the goal is consensus. When really the goal is, “Let’s hash it out together. We may not all agree. One person will be the decision-maker, and then we will all commit to it.” If you don’t clarify what kind of decision this is, then groups really struggle because their expectations are not set correctly.

When I’m leading through a tough decision, I try to say, at the outset, “I want all of your opinions, but I’m going to be the one who ultimately makes the decision.” Or in some cases, I will say, “I don’t know if I’m the right decision-maker. I need help exploring what the decision vectors are, and I need all of your help. And then I will let you know how we’re going to make the decision once we’ve talked about it.” If you don’t give people that guidance, which is I think a common mistake, you’re likely to run into trouble.

Bi-weekly or weekly 1:1s. We’ll try to keep the times consistent so you can plan. I’m a big fan of a joint 1:1 doc to track our agendas, actions, goals, and updates.

You can keep the board small in multiple ways: one is, don’t give up more than one board seat per round.

Yeah, I’ve seen that situation actually kill companies. My solution to most board problems is going to be highly unacceptable to the venture community, but we don’t give out permanent board seats. We would never ever give out a board seat that we could not remove. And in AngelList and my company, that’s what I’ve done. The only board seats I’ve given out are ones that can be removed; there’s no such thing as a permanent board seat.

The way my company is structured is that everyone can be removed, including me. Then it is actually a very consistent moral argument that I can make, which is saying, “Hey guys, you can remove me as well if I’m out of line.” Nobody is safe, and that forces everybody to behave.

Reference checks are often the clearest signal on a candidate. You should referencecheck everyone.

Things that impact the ability of the recruiter to be effective include: Brand of the startup with candidates. Strength of the hiring manager and executive team as recruiters. If they are active and engaged, it makes recruiting run smoother and will help to source and close more candidates. Breadth of network of the employees at the company.

Send out a welcome letter Send a welcome letter to the new employee—and cc all the teams that person will be working with closely. The letter will explain the person, their role, who they will report to, what their goals are for the quarter, as well as potentially one interesting fact about them they are willing to share. The idea is to ensure each new person has a clear role and responsibilities, and that the rest of the organization is aware of these things. The interesting fact creates an ice breaker for their coworkers to be able to start a conversation with the new hire.

Buddy system High-growth companies tend to have their own jargon, internal tools, and random processes that are unique to them. Pair a new hire with a “buddy”—someone who is not in the reporting chain of command with them who can take them to lunch, introduce them to people, and importantly answer any “stupid” questions they may have. Buddies tend be paired for one to three months.

Set goals Each manager can set 30-, 60-, and 90-day goals for new employees. This gives a sense of direction, context, and structure for the new employee. It also emphasizes what is important to get done and that individual’s priorities.

A common sign that an old-timer will work out is their eventual acceptance that their role and influence at the company will shrink in the short- to medium-term as the team scales, but that it will expand with time as they continue to learn and the company continues to scale.

You just have to relentlessly say, “This is what we’re doing, this is why, and this is how we’re going to do it.” And that part—the communication and the evangelizing of the company vision and goals—is time-wise by far the biggest part of the job.

If the CEO disconnects from the product, that’s usually bad. And that is something that you see happen to varying degrees as a company scales. There are a lot of CEOs that say, “Well, I want to just go think about strategy all the time. I’m really tired of managing people.” Because managing people is hard.

No one is naturally good at letting go of something they care deeply about. So, I think you just have to give yourself permission to screw it up for a while. You can internalize that it’s important to delegate and give people authority, but you won’t do it well the first few times. Just continually try to get better at it.

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?

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.

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.

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.

A lot of times, at the board level, a conversation will start with, “Do we have the right CEO?” And especially for early founders, you ask yourself that question. “Am I going to be a good CEO? Am I going to be able to scale the company?” But when you ask that kind of question, you end up defining what a CEO is in unnecessarily strict terms. Whereas, actually, the most important job of a CEO is just to make sure that the company succeeds. So if there’s a whole chunk of work that you’re not good at, you don’t personally have to be good at it to be a good CEO. You have to make sure the company is good at it to be a good CEO.

you need a COO to deeply integrate into the company as quickly as possible. In the first 30 days, they should go and interview every single employee that they can get their hands on—about what’s working, what’s not working, what they want to see improved, where they want to take the business. The key is to just get very, very integrated into the culture, and get to know the individuals involved, in the first 30 to 60 days.

Raw charisma is drastically overrated by technical founders. Don’t be fooled just because someone is friendly and charming.

It’s a matrix organization; it’s not a GM organization.

Prioritizing it, for us, is mostly determined by the growth road map, but then sometimes it’s about productivity of people. And sometimes it’s about pure happiness. If there’s a tipping point where people are really unhappy about something, obviously we do take that into consideration.

we communicate many, many, many times what the overall strategy of the company is, and the corporate goals.

Finance has given classes in ROI calculations to engineers. We have a lot of finance classes, essentially, that we’ve organized for the engineers to understand how to commercially evaluate these things.

You should never start with the cheap solution of building a whole team of young engineers. Whereas there are other functions where you can potentially get away with it.

The only place where I think engineers might get into trouble is on the marketing/finance side. I see that a lot. There’s this very important issue of getting your business model straight. And just as businesspeople have trouble figuring out who’s a great engineer, I think sometimes engineers have a really hard time figuring out who’s a good business person. It’s just this crazy divide unfortunately. There are not that many people who really understand both the financial side and the engineering side.

Communicate to the team that as your company grows quickly, things will shift around. Make it clear that it is normal for that to happen—it’s a sign of your success—and that other companies that grow fast do the same thing.

re-orgs will occur at the company frequently.

Re-orgs should never be open conversations with the whole company (or a functional area) about what form the new organization structure should take. This only opens you up to lobbying, internal politicking, and land grabbing. It also prolongs the angst—re-orgs should happen swiftly and with as little churn as possible.

4. Announce and implement the re-org soup-to-nuts in 24 hours. Once you have decided what form the new organization will take, discuss it with your reports in their 1:1s. Your executives should have a clear plan for how and when to communicate the changes to their team members. If there are key people deeply affected or likely to be unhappy with the change, you or one of your reports can meet with them either right before or right after the announcement to hear them out and re-affirm the logic for the changes. You should never drag out a re-org or pre-announce it. Try not to announce “this week we will re-organize product, and next month we will change engineering.” If possible, all elements of the re-org need to be communicated and implemented simultaneously. If you pre-announce a portion of the re-org, that team will not get any work done until the re-org happens. Instead, there will be hushed conversations in conference rooms full of gossip and speculation, crazy rumor-mongering, and executive lobbying.