Last week was the end of my 8.5 year tenure at LiveRamp. Over that time period, I started as a summer intern, led product management as the business grew from $0 to $40 mm, and served as CEO as we scaled from $40 mm to ~$200 mm in revenue and ~400 employees (I stepped out of the CEO role 6 months ago, and have been transitioning out to found Datavant).
I leave remarkably proud of what we built, and very bullish on LiveRamp’s future. As I’ve reflected in my last few days, I wanted to list out some of the key lessons I’ve learned that have made me a better CEO. I’ve tried to focus this article on the non-obvious lessons, and have also avoided some of the lessons covered in other sources of management advice that I enjoy (such as lessons contained here, here, here, here, here, here, and here).
In no particular order, here are the lessons:
1. Define your own CEO job description
2. HR is the most undervalued function, and the most important one
3. Customer service is the second most undervalued function
4. Implement “management by random sampling”
5. Avoid the plague of headcount growth.
6. Retain control over who you hire.
7. Keep your best frienemies close.
8. Fight complacency — “no broken windows.”
9. Maintain a sense of “company exceptionalism”
Lesson #1: Define your own CEO job description. The CEO job description is one of the most open-ended charters in any company, but it’s one of the jobs that comes with the most preconceptions. Different CEOs are extremely different — some CEOs are great sellers, some product visionaries, some charismatic cheerleaders, some great at leading from behind. The key to being a great CEO is to create the role in a way that plays to your own strengths, and figure out how to hire and empower around your weaknesses.
I am not what people typically imagine when they imagine a CEO. I’m an introvert who hates public speaking. While there are a few areas that I’m best-in-the-company at, I’d be mediocre as a sales leader, mediocre as a marketing leader, and poor in several other functions. So I hired around these weaknesses, and I empowered an exceptional GTM leader and charismatic #2 and #3 leaders. I designed my role to spend as much of my time & energy as possible focused on the areas I had the greatest comparative advantage, and built the team to compensate for my weaknesses. This maximized the impact I could have on LiveRamp.
A corollary is that a great CEO must closely guard their time. Everyone pulls on the CEO’s time: time with clients, key deals, key internal meetings, facetime with every team, engaging investors, tracking the business’s performance, and closing candidates are all important things to get done. I had to learn to aggressively say no, aggressively delegate meeting attendance, and spend time where it could be most impactful. My rule of thumb was to always have at least 30% of my time that was “discretionary.” I budgeted 70% of the time towards understanding the key activities of the business and handling things that truly required me, but then for the remaining 30% I prioritized 1–3 key projects at a time that I’d focus on proactively. Whenever I fell below 30% discretionary time, it was bad for the company in the long run, and I knew I’d need to say no to more and have others step up.
Lesson #2: HR is most undervalued function, and the most important one. In a hypergrowth business, the number one thing a CEO spends time on is HR: thinking about how to scale recruiting, how to scale culture, how to design the org structure for the future, and more. This is the most important part of the CEO job, and the hardest.
Unfortunately, most CEOs don’t think of their HR leadership hire at the same level of importance as their Chief Product Officer or Chief Revenue Officer, and most HR roles (and most HR professionals) become compliance-driven. This is a huge mistake: an incredible head of HR gives the CEO tremendous leverage in designing the enterprise, the culture, and the team that reflects the CEO’s vision — and is the most important executive during hypergrowth.
Lesson #3: Customer service is the second most undervalued function. Most SAAS companies look at technical account management/implementation/customer service as cost centers getting in the way of higher gross margins, and attempt to make this function as low-cost as possible.
This is a mistake. At LiveRamp, we generally tried to hire people overqualified for a traditional customer implementation role (we often promoted it as an entry point for high-potential generalists). A great customer implementation team should be focused on automating itself out of jobs, and thinking about the system, not just the specific customers.
More importantly, our customer implementation team was the source of every major product idea we launched while I was at LiveRamp. Most “new” products would start as a feature request from a specific client, and a Product Implementation Manager would see it, realize the generalizability of the idea, and champion investing in it (and usually figure out some hacky way to get it delivered to the client that requested). The key is to do this for just the ideas that actually ought to be productized; so a great Customer Implementation Manger would understand the frontier of our product, our overall strategy, and clients extremely well — and would know what to champion to the Product team.
Not only did great products come out of this approach, but this approach worked financially; by hiring superstars in customer service, we had >120% net retention rates most years, and a typical customer implementation manager automated enough that they could handle a several million dollar book of business, keeping our margins high.
Lesson #4: Implement “management by random sampling.” In general, as a company scales, executives inevitably lose touch with what’s actually happening on the front lines, and have layers of management/“noise” between them and reality. Executives that can’t handle being so removed usually become micromanagers, and are unable to scale (and people don’t like working with them).
I learned a hack that’s a middle ground approach that I call “Management by Random Sampling.” Within the company, at any given time, I would try to track a small number of random projects — non-exceptional customers that are about average ACV, routine engineering projects, etc. I would go extremely deep on understanding these — what does the customer want, what are the bottlenecks, how quickly are we communicating, does the sales rep explain the product well, etc. But I would keep my involvement passive (watching passively on JIRA, on Salesforce (where all of our external emails were tracked), on internal aliases, etc.).
By picking a random sample (and not just focusing on the big issues facing the company, which is a highly skewed sample), I could be comfortable that if things were going well in the sample, they were going well in general. Over time, this approach helped me to quickly detect patterns of internal communication problems, team misalignment, confusing product collateral, legal bottlenecks, and several other issues facing the team.
Lesson #5: Avoid the plague of headcount growth. No matter how well run a company is, a 500-person company rarely has more than double the output of a 50-person company; the effectiveness per person drops precipitously. As you add people, you add communication overhead, internal misalignment, and potentially politics (and obviously, you add cost).
The first way to avoid hiring is to hire superstars; even if they cost more, having a team of people who can accomplish more per capita means you can keep headcount down, and avoid needing to scale too quickly. You should also instill in your culture that hiring is a sign of failure, it means that you didn’t automate sufficiently; teams should obsess with leverage and how to make each person dramatically more effective.
The other approach I found effective was to actually split the business as it became too large, even at the cost of internal synergies. During its peak growth stage, HP had a philosophy that it would always divide a business in half once it crossed a particular threshold of revenue; this approach is brilliant as it turns a large company into a confederation of high-efficiency startups (I tried to do this with limiting business units to 100 employees). Giving up the internal synergies is hard, but it is well worth it to keep things efficient.
Lesson #6: Retain control over who you hire. One of the last responsibilities I refused to delegate as LiveRamp scaled was control over hiring; until we were >200 employees, I personally interviewed every non-engineering hire we made (the engineering leadership team and CTO were pickier than I was, so I stepped out of engineering recruiting earlier). Once we crossed 200 employees, I still approved every offer package personally, even as we hired 100+ people in a year.
My personal involvement did a few things. First, I was adamant against “compromise hires” — usually, the hiring manager has an incentive to hire quickly. Second, seeing how others on the team scored candidates helped me calibrate where in the organization leaders were good at assessing talent, and where they weren’t. Finally, my involvement in the hiring process and my reputation as a tough grader signaled to others that they should care about hiring and that they should also be tough graders.
Lesson #7: Keep your best frienemies close. I hold an extremely “positive sum” approach to business development; at LiveRamp, we genuinely believed that we could sit down with anyone and find a win-win way to grow.
As we’ve grown, many of our best partnerships have been with companies that were somewhat competitive with us. For most of these “competitors,” even though 20% of their business competed with 20% of our business, the remaining 80% could be extremely complimentary on both sides. We would openly discuss where we were aspiring to go, where they were aspiring to go, and how we could each help each other on the remaining 80%…and would be fine cooperating while still competing on some parts of the business.
As a corollary, we also decided early on that we did not care much about account control; if we could scale more effectively by handing account control to a partner (while retaining data flow control and product control), that was fine for us.
This willingness and bias towards partnership helped us quickly build scale in our early days, and we ultimately found that channel conflict was more hypothetical than real as a market was growing.
Lesson #8: Fight complacency — “no broken windows.” During William Bratton’s time as police commissioner in NY, he popularized the “broken windows theory” of policing; that is, if the police force doesn’t crack down on small crimes (represented by “broken windows”), it creates an environment of lawlessness that ultimately leads to more big crimes.
While this is very debatable as a matter of criminology, it is 100% true in company cultures: if you allow the company to become complacent on the small things, it gradually becomes complacent on the big things.
Making sure the organization sweats the details of the small things is critical; it creates a culture of feedback where people comment on (and fix) mistakes impacting clients, bugs, bad formatting, office layout, internal misalignment, and internal processes; it keeps the organization from ever accepting complacency.
Lesson #9: Always maintain a sense of “company exceptionalism.” As CEO, I considered it my job to be 2/3 extremely optimistic and 1/3 extremely paranoid; any other balance would either mean too much hubris or too much fear to be successful.
The optimism had to come from a sense that we’re truly different than everyone else, or “company exceptionalism;” To have a sense of exceptionalism, a company should be very self-aware of what makes it different, and embrace and double-down on those traits. The company culture should celebrate “what makes us exceptional.” It is an exception to the typical rules of business, the rules that say the odds of success on big bets are low, you can only grow at a certain pace, etc. This sense of exceptionalism is what fuels greatness.
As you get bigger, companies start to believe less in their exceptionalism; single-product success eliminates the urgency to continue making more big bets, and gravity can pull an organization back to reality. A key part of the CEO’s job is to keep pushing the exceptionalism — and maintain both the narrative and the reality of being different.
These are just a few of the big takeaways I had, but some that have shaped my outlook as I double down on Datavant. Thank you to the many, many people who have shaped the growth of LiveRamp and my experiences there. The journey of the last 8.5 years was truly incredible.