by Elad Gil
Early on, when we were scaling Box at 20 or 30 employees, we came to this question. And I asked myself, “Do I look like these other CEOs that have scaled big companies?” A lot of times it didn’t feel like I was like them. I focus on a very different set of things. I don’t really like all the stuff that they like to do. But I ended up realizing that it was actually just my job to make sure that we got all of the problems solved, not that I had to personally solve them all.
I end up seeing a lot of companies make this mistake—and it’s obviously happening much less these days than in the ’90s and early 2000s—where you shoot the founder and then bring in an experienced CEO. Imagine all the better outcomes we would have if those companies actually kept their founders focused on what they’re great at, and then added counterparts to help them build the business.
Elad: So you see the COO as a counterpart to the CEO. How can leaders go about defining the role in concrete terms?
Aaron: I think COO is probably one of the most generalized jobs, but ultimately one of the most specific to the individual and to the company. Unlike CMO or head of product or whatever, where generally you know what’s in their purview, COO is highly variable. If you found five companies with COOs, the job descriptions of those COOs, or the things that they were responsible for, would be different by each company. The way that Sheryl Sandberg operates at Facebook is likely different than how the COO at Yelp operates, which is different than how the COO at Box operates. So first, I would just dispel the notion that COO is one particular type of job or type of function.
I ended up bringing on a COO because I wanted somebody who had a depth of experience in scaling up companies and organizations and working in large environments, to be a sort of counterpart as we were building the business. I cofounded the company with a friend from college, and neither of us had any kind of real scale experience. So by bringing in a COO who had managed a thousand- or multi-thousand-person organization and had managed multiple hundreds of millions of dollars of revenue for a very large business, we could bring on experience in all the sorts of disciplines we were looking for.
In our particular situation, the COO that we brought on was formerly a general manager of a large division of Intuit, and a CEO prior to that. So he had a lot of experience owning all the major functions of an organization. He had a particular interest in things like talent development, organization design, and scaling up of organizations—so a very people-oriented COO.
In general, we created a structure where he really owned building up the company and really owned building up the individual organizations. And then my job was focusing on strategy and where we want to go over the long run. I stayed highly tuned toward things like the product and the product strategy.
Elad: That makes a lot of sense. It seems like many people claim that the COO role is sales and the CEO handles product and engineering. But there are so many counterexamples. So that really resonates with what you’re saying: each company is different.
Aaron: Yeah, I think that my best description of why and when to get a COO is this: If you as the CEO are uniquely strong in a couple of areas, and you want to supplement or complement those areas—if you need to fill the gaps—consider a COO.
No matter what, your business needs to have a well-running set of organizations. And if you’re not interested in managing those organizations, or you don’t have the bandwidth, then you need some kind of counterpart to help you.
If you’re a very sales-oriented founder, then you might get somebody to augment on the product side. If you’re a very product-oriented founder, you might get somebody to augment on the sales side. If you’re a very sales- and product-oriented founder, you might get somebody to augment you on the pure operational infrastructure side. There are so many different flavors of what this means. In most cases, it’s actually just as important to understand and be clear about what the CEO does as it is to understand what the COO does.
Elad: How early do you think people should consider a COO? I think Stripe brought on Billy Alvarado when they were, I’m guessing, eight or ten people. But I know a lot of companies wait until they’re 100, 200, 300. When is the right time?
Aaron: I think it doesn’t make sense until there’s some sort of inflection from a growth standpoint. The idea is not to hand a COO random busywork from around the company. Instead, consider some key criteria: Are you scaling at a rate where your own time is now spent primarily on pure operational activities—like hiring and building up goal-setting structures and performance reviews and helping people with organizational issues and process? Is that pulling you away from things like your product strategy and product design or working with customers, for instance? You have to solve all those problems either way, and at some point in the inflection curve, it just makes more sense to get a partner to help you with the operational stuff.
In some companies, that could be when you hire your twelfth employee. In some companies, if you’re growing at a steady rate, instead of instantaneously, that could be once you reach a couple hundred people. You might not need it very early on. So I think a lot of it is the rate of change of the business. Add the COO role when the rate of change is reaching some kind of escape velocity.
“I ended up realizing that it was actually just my job to make sure that we got all of the problems solved, not that I had to personally solve them all.”
—Aaron Levie
Elad: What background did you want in a COO? You mentioned that for other roles it’s very clear—if you’re hiring, say, a VP of product, you’re looking for very specific experience. But what would you recommend people look for in particular in COO candidates?
Aaron: Again, you have to first ask yourself either: What are you not good at, or: Where are you spending time that you don’t want to be spending time? And you have to make sure that there’s a surface area that is large enough for a COO. If you’re just finding yourself in a lot of legal conversations or finance conversations that you don’t want to be in, then really what you need is a general counsel or an attorney, or a CFO or head of finance.
So the first thing is to actually identify the areas where you’re going to want a partner in helping you scale. From there—and this is one of those things where you should go on a mountain and light some candles and really tap into your inner self—honestly answer some key questions. What are you really good at? What do you want to get better at? And what things aren’t going to be a competitive advantage for you to personally get good at? Because you can’t be great at everything.
Elad: What did your list look like? How did you get started with the search?
Aaron: For me, the list was a bunch of more generalizable things, like management and organization development and adding processes to the company. Those were things that I was uniquely bad at. As soon as we were at 40 or 50 people and we decided we really wanted to scale the business, we realized that it was going to require a different kind of mindset and mentality than what I had and what my cofounder had.
In our COO search, we actually meandered a little bit. We thought maybe we wanted a chief revenue officer, but then we realized that we had, again, a more generalizable issue around a lot of the operational stuff. So ultimately we came up with a job spec that was about getting somebody in here that had seen a couple thousand people, had seen a few hundred million dollars in revenue. From a job spec standpoint, it was: Has seen scale, has managed scale, has driven very large revenue targets, has been in a somewhat operationally complex, or at least multifaceted, environment—so sales, marketing, customer success, product, engineering, etc. We wanted someone who knew their way around lots of different parts of the business.
Once we found somebody with all that, and who was a great cultural fit, the next question was, can I actually get along with them? Can we divide and conquer in a very collaborative way? Being a COO at times can be a very thankless job, where you have to leave your ego aside to some extent. So you have to make sure that you have a very c
ollaborative structure, and that each person deeply respects the other person. You have to make sure it’s somebody that you do want to spend a lot of time with—because you’re going to be in the trenches with that person.
When you add up all those factors, it becomes one of those things where there’s a whole bunch of people you’ll have big question marks around, and then there might be one or two people where you think, “Okay, this really fits.” And you want to wait for the person that really fits.
Elad: How did you figure out who was the right person for you? What was your hiring and interviewing process?
Aaron: We had a very roundabout way of doing this, so I don’t think that my experience is a lesson that can be replicated easily. We had an advisor that was introduced to us by a board member. Through working with that advisor, I don’t know, a dozen or so times, I got the sense that there was a whole world of management and building companies out there that I had never really imagined. Because I was a very product-focused person. I thought, you just build a product and everything falls into place from there. And just the vocabulary this advisor was using and the concepts that he was imparting on me—it was like learning a completely new trade. And I realized that we really needed to get better at this kind of stuff.
And so we actually brought this advisor, Dan Levin, onto our board. Then we kicked off a search for a CRO/COO, and after three or four months of that search, we’d met a whole bunch of people. And every single time, we realized that I didn’t get along as well with any of those people as I did with Dan.
I ultimately spent two or three months just trying to convince him to join. So it was probably six-plus months, or maybe nine months, from meeting him to having him join as COO. But it was hundreds of hours of interactions between me and executives and the board and him.
Elad: Are there things that you think CEOs should never give away? For example, are there certain things at all-hands meetings that only the CEO should ever talk about? Should the CEO always run board meetings?
Aaron: Again, this totally comes down to what the CEO is good at versus just what the COO is good at. I don’t think there’s any rule here that I’ve seen. Our COO runs the executive staff meeting. I think it’s probably indistinguishable who runs our board meeting between me, our COO, and our CFO.
The most important thing is just to make it clear who owns what. That’s a whole element of discipline around all of this: What jobs and what decisions do you separate between the two functions? And how do you make that clear to as many people as possible in the organization so it’s not confusing?
For us, we basically said, “If you have a people-related issue, if you have an organization-related issue, if you have a process issue, you go to Dan.” And in fact, from a functional reporting standpoint, most of the executive team, except for my cofounder, actually report to Dan. On the other hand, if you have a major strategic thing, or if it’s product-related or if it’s brand-related, then I’m going to own that and get involved in that.
So this means that at times, as an executive at Box, you have to go to one person about one thing and a different person about a different thing. But it means that, when it comes to doing your job, you are hopefully getting a non-diluted experience. Dan is far, far, far better at all the things that he is good at than I am, and I have a fairly long-term view around the strategy side that has ended up being helpful as well.
Elad: Is there anything specific that you felt was important that you did when you onboarded Dan as COO, or anything that you wish you’d done differently?
“You have to make sure it’s somebody that you do want to spend a lot of time with—because you’re going to be in the trenches with that person.”
—Aaron Levie
Aaron: This was a long time ago! But I’ll say something kind of generic that would apply to almost any executive: 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.
Having some early wins is certainly helpful too. So somewhere in the 60- to 90-day time period, the new COO should implement a few things that make everybody go, “Oh wow, this is cool. This is helping. We now have this new process that makes us perform a lot better and has reduced wasted effort in the wrong areas.” Having some early wins to show immediate value, and getting to know the company and the culture—that’s all super important early on.
But again, this stuff breaks down because of swim lanes more than anything. The key is just to be crystal clear about what each person is going to own and ensure that you have a process for resolving things when there is conflict or overlap or a gray area. Because there will always be some gray areas—it’s not like it works perfectly. But starting with as much clarity as possible is incredibly important.
Elad: Last question before we wrap up: We’ve been proceeding from the assumption that a founder wants to stay on as CEO. But when should founders acknowledge that they don’t want to keep going, or that they shouldn’t keep going?
Aaron: My thing is this: Founders can get burned out for two reasons—I mean, probably a hundred reasons, but let’s just say there’s two for now. One reason is that you’re tired. You just aren’t as passionate about the problem space anymore, and you want to go off and do something completely different or take some time off. That’s scenario one, and realistically it probably happens just as much as scenario two.
But I think sometimes it can get confused with scenario two, which is when you are doing a set of things that you no longer like to do, and you have not found a way to get them off your plate. You haven’t found a way to get back to focusing on the thing that you are uniquely passionate about. If I spent 90% of my time in performance-review meetings and organization meetings and compensation-decision meetings, I would be burned out and say, “This is not what I want to do.”
The key is to not conflate what you are doing on a daily basis with falling out of passion with the overall mission. But a lot of times you end up doing that if you don’t have any way to complement your skills.
So my recommendation to anyone who’s running into that situation would be to try a COO first. See if the desire to stop has to do with dreading your day-to-day activities, or if it’s actually where you are in life.
Executive titles and pragmatism
Companies tend to have two philosophies on titles: The Google approach, which Facebook copied, was to give low titles to people as they joined. People who were VPs at Yahoo! or eBay joined Google as directors or manager: while people who were VPs at small startups may have joined as individual contributors or junior managers. The alternative approach is to give everyone big titles to partially compensate them for the risk of joining an early-stage company. Neither approach is right or wrong, and both have trade-offs; you should just choose one and apply it consistently. I personally favor the “try to keep titles low” approach for as long as possible, as I think it decreases hierarchy and prevents people from valuing a VP’s opinions more than a non-VP’s.
When hiring executives, you’ll find that candidates may or may not be willing to capitulate on title. By joining a startup, they could be leaving millions of dollars on the table and will argue that a step up in title helps to compensate for the risk (e.g., a director from Google will want to join your startup as a VP or maybe COO, CMO, etc.).
One way to deal with this is to negotiate a vague title—e.g, “Head of Sales” versus “VP of Sales,” or “General Manager” rather than “COO”—or agree to defer the title choice to a later date. You can also agree that if the person performs at the level of the title they want, then when you roll out titles for the broader team they will get that VP or COO role.
Finally, remember
that it is difficult to hire someone above a person with a CXO title. That is, you can hire an SVP, EVP, or COO above a VP of Sales. But if you make someone who should have been VP of sales the COO, you are likely to have to let her go (or offer a demotion that will likely lead her to quit) if you want to hire above her one day. Similarly, a director-level title is easier to hire over than a VP-level one, although both are doable.
Firing executives
It is always painful to let people go. When you fire an executive, it can be hard not only on you and the executive but also on his entire organization.38 There are some steps you can take, though, to minimize the uncertainty and confusion that come along with these difficult decisions.
Before you let an executive go, you should make sure to have the following in place:
1. Discuss it with your board. Your board of directors will want to be aware of any key executive changes before you make them. It is possible one or more of your board members has worked with the executive before, or referred them to you.
It is best to call each board member 1:1, explain to them why you want to make the change and how you plan to do it (including a potential transition plan), and if needed the details of any severance package. Once you have spoken to people 1:1 you can do a follow up board call if needed in the case that active discussion is necessary.
