What Does a Startup COO Actually Do?

July 15, 2015

Titles shouldn’t matter, but unfortunately everyone has to have one — including me. Today, I am the COO at Mattermark. The only consensus on what a COO does is that there is no consensus, especially at startups (see articles at the bottom). The role varies from company to company, some don’t have the role, and the jury is out on whether startups need one.

I believe the job of a founding COO is to do the most important things the company hasn’t hired for yet, hire or delegate to someone in your place, and move on.

Founding COOs need to be zero-to-one people who are good at and comfortable with firing themselves.

How my role as COO evolved

I wasn’t always the COO. Mattermark is the same legal entity as Referly, but as Danielle says, “Mattermark wasn’t a pivot.” I was not a co-founder of Referly. Danielle and Kevin actually hired me after I shut down my startup, LaunchGram. I was an individual contributor hired when Referly was gasping for air, and my job description was essentially, “go on missions for Danielle.”

I remember the lunch where Danielle told me she and Kevin had decided to shut down Referly, and I wasn’t totally sure I had a place in their post-Referly company. Danielle explained to me that they had no idea what they would do next, but that I was welcome to join them and continue going on missions.

So that’s what I did, and it’s not that different from what I do now.

When we started Mattermark, Danielle blogged a ton and talked to potential investors, I hand-gathered a boatload of data, and Kevin helped me turn my process into software. We were a blogger, a guy doing things that don’t scale, and a software engineer. When we were on the eve of launching Mattermark, I had a bit of an epiphany.

I asked Danielle and Kevin, “Am I a co-founder of this new thing? Because I feel like one.” They reassured me I was. Going forward, Danielle was our CEO, Kevin was our CTO, and I didn’t really feel a need to have a title more defined than “Co-Founder.”

Mission #1 : Start our data team

In the early days we got our data by hand. We had an advantage over incumbents and later up-and-comers because we were willing to do work that didn’t scale to deliver value for our customers.

Early days. Collecting data on Benchmark's portfolio.

So the first team we built at Mattermark was a team of data analysts who filled a forest of spreadsheets with data on companies in VC firm portfolios. I led this team in my first functional owner role, and we decided to change my title to “Head of Research & Data.”

As we built repeatable processes to get certain data points — we call it the “human algorithm — we realized I had taken the team from 0–1. It was time to hire a much more qualified Head of Data.

Mission #2 : Figure out sales

In January 2014, we had just tried to raise our Series A, and got feedback that we needed to prove out some other customer segments. Our sales were entirely inbound at the time, and we were getting a lot of non-VC customers signing up from investment bankers to salespeople to real estate brokers.

I set out to (1) increase revenue and (2) determine whether we were actually ready to sell to any of these other segments.

I was good at this mission, but I didn’t want to do eight demos a day and carry a quota forever. This was hard both for me and the company. By firing myself from sales, I was essentially firing one of our most productive sales reps.

Takeaway: As a founding COO, you will have to pinch hit and do things you don’t like. Eventually you have to find someone else to do those things though, or you’ll burn out.

Early on in my second mission, at one of our company all-hands meetings, I remember when one of our engineers asked, “Andy, I know you’re working really hard, but what is it that you do?” We all laughed, and I explained the things I had been working on.

I felt like an imposter though, because my latest mission was as an individual contributor. Co-founders are supposed to manage teams and run the business, right? I believed I had to directly manage people to be a leader. I was asking myself what I was going to lead. Ultimately, I knew an opportunity would present itself and that as long as I focused on driving value for our customers and our business, I was in the right place.

Takeaway: Changing roles often is hard enough for the person doing it, but it can be confusing to the rest of the team. Make sure to clearly articulate your new responsibilities .

Mission #3 : Becoming “COO”

My third mission was less well defined because, well, it was about five missions at once. This fall, we needed to get a new office, figure out recruiting, start a customer support & success team, and turn our startup into a professionally managed company (see legal, accounting, HR, and all the fun stuff).

At this time, I had the choice to go with “VP of Facilities, Recruiting, Customer Support & Success, Finance, & HR,” or something more catch-all like COO.

One of our investors offered a third option, “VP of Shit.”

I decided to go with COO.

The Future

In 2015, we hired an outstanding Customer Success Manager, Manali Karmarkar. That team is running like greased lightning now. We moved away from a centralized recruiting function with internal recruiters to a model that distributes the load amongst hiring managers. Finally, we hired a VP of Finance, Emad Khan, to take over finance, HR, and facilities.

Keith Rabois says “you’re constantly fixing things,” which you are. You are also constantly solving problems you haven’t solved before. Problem solving in an ambiguous environment should be a founding COO’s core skill.

Going forward, my job will be the same as it’s always been: find the most important things the company hasn’t hired for yet, hire or delegate to someone in my place, and move on.