Executive Function: Building Systems That Can Make Decisions Without You

When you think about executives at scale-up startups, what’s the difference between someone who’s very good and someone who’s extraordinary?

Some of what I would say is — and this is something that Stripe valued, and I think a lot more companies are generally valuing — you have to be able to operate simultaneously at 30,000 feet and ground level. From a 30,000-foot perspective, you have to develop a multi-year point of view on where you ought to go and how to get there. That feels harder than ever right now. With AI, you just don’t really have a sense for whether a two-year plan is reasonable.

And at the same time, you are still hardcore building a company. There are all sorts of operations that are not yet operational, and there can be a pretty big gap between that 30,000-foot view and what needs to be true to realize it. Not everybody at your company has had the experience before to know how to turn that North Star into an hour-by-hour, minute-by-minute operating reality. So you’re in some cases building that yourself. In other cases, coaching somebody who has the will but not yet the skill. And in other cases, being able to delegate more freely to somebody who has done it before and can bring that vision to life.

What outcome is an executive actually responsible for? As an IC it’s very clear — you’re a PM, you’re shipping a product. How do you think about results as a COO?

You’re thinking about things on multiple time horizons. I am accountable to specific results within a year — but an executive is meant to be long-tenured. My career has been marked by basically two nearly decade-long tenures. Executives have decade-long stints because you are able to replicably produce results year after year. That means the bulk of your team is executing for in-year results, a set of leaders are thinking about next year, and you, as a leader, need to be thinking even further out. How is the competitive dynamic changing? How is the market around me changing?

The other thing you’re doing is looking more broadly. Any really strong executive typically thinks more as a general manager with depth in their own area — but also an ability to make trade-offs that might not be in their own self-interest because it’s going to set the company up for success in the long run.

What tends to keep star talent from ascending to the C-suite? They’re incredible VPs — they never become generational COOs or CROs. Is there a common thread?

I do think it’s an ability to transition into systems thinking. If you move from IC to operator to leader, it’s not about you anymore. It’s about: can you get a group of people to execute the way you, as an IC, would have done a thing?

One of the biggest leaps in leadership to make is going from a frontline manager to a second-line manager. As a frontline manager, you’ve got seven direct reports and you’re pretty hands-on. When you get to manager of managers, you can’t be in every deal, every product review. You now have to design a system. You have to understand what metrics you need to be watching, what your leading indicators are, what forums you need to review, who your key people are, how you keep a pulse on what’s actually going on through them. Where people get stuck is they try to scale via what has made them successful so far — and basically just continuously become a super IC rather than empowering the other people around them.

If someone is effective at their current stage and their goal is not to be layered or managed out as the company grows, what should they be doing?

You need to work yourself out of a job. You’re comfortable doing the job you have become good at. That is no longer the job. The people who survive scaling are the ones who, as they get better and the team gets better, figure out what the next job is — and start working on that. If you don’t, someone will hire above you because you’ve become too comfortable in the role you own.

Can you walk through your executive interview process? What makes it rigorous?

My first interview is mostly bidirectional — learning about the candidate’s background, going deep into what they’ve done, and selling them on the role. But I lay out very clearly: “These are the reasons I’m hiring for this role. These are the specific issues you will need to come and resolve.” By putting those things on the table in the first meeting, you very quickly get a sense for whether that person gets motivated by solving those things or not.

My second call: I send a prompt beforehand asking them to go deep on those four areas — what have you done that’s similar at prior companies, and how might you start to tackle it here? The third stage fans out across the leadership team. Then there’s a 90-minute workshop with three prompts that align to what they would be expected to do in their first 90 days. That workshop is where you really see work product rather than answers to interview questions. After the workshop, there’s a one-on-one with our CEO — because any executive reporting to me has to be someone he also trusts and would go directly to.

What’s the common thread in executive hires that don’t work out?

I think a lot of executives over-pattern-match and do not have the intellectual curiosity to figure out what’s different. They import their playbook from the last company wholesale. What worked at a 200-person company doesn’t necessarily work at a 600-person company, and what worked at a B2B company might not work at a developer-first company. The executives who stay and thrive are the ones who are genuinely curious about what makes this situation different.

The other failure mode is not building the right bench beneath you. You can execute brilliantly for the first 18 months and then hit a wall because you haven’t developed the leadership layer that would let you go higher. The ceiling isn’t the executive themselves — it’s what they built around them.

You mention driver trees as a core approach to metrics. What’s the philosophy there?

A driver tree starts at total revenue — or whatever your top-line metric is — and breaks it down into the hundred different nodes that actually drive that outcome. The goal is that throughout the course of the year, if revenue is green or some other color, you know exactly what’s working that you can lean into, or what’s not working that you can course-correct.

The mistake most teams make is they have a small number of headline metrics and they don’t know why those metrics are moving. A driver tree gives you diagnostic capability. When something goes wrong, you can trace it. When something goes right, you can replicate it. I try to get that in place anywhere I land — it’s a biggie.

What was the most valuable performance review feedback you’ve ever received?

One piece of feedback I received was: “Cross-functionally, when you don’t think somebody is good at their job, they’re dead to you.”

I couldn’t unhear that. And when I really sat with it, it was true. When I made a judgment that someone wasn’t performing, I mentally moved on from the relationship. I stopped investing in them. I would work around them rather than working with them. That’s not good leadership — it’s a pattern I had to actively break. The most important feedback is often the feedback that shows you a blind spot you didn’t know you had.