Everything in this series so far has been about building inertia — a cadence, long-term goals, processes that keep a growing team moving fast in one direction. That’s the point of it. But here’s the trap: the same machine that lets you scale is very good at resisting change. The moment you want to launch a new product, go after a new buyer, or enter a new market, you’ll feel your own systems push back.
Which gets at how I actually think about the job. I tell people my role as CEO is about 50% solving problems and 50% creating the right ones. The first half is most of what this series covered — keeping the business running well. This post is the second half: deliberately disrupting the thing you just spent two years stabilizing, because standing still is how scaleups die.
The framework I use for it is Geoffrey Moore’s Zone Offense. It sorts the work of a company into four zones, but the split that matters here is simple: most of your business lives in the Performance and Productivity zones — the core operations, the established products and markets, the stuff you run with the cadence and tools we’ve been talking about. Those zones mostly take care of themselves once they’re humming. The two that need you personally are Incubation and Transformation.
Incubation: planting seeds
Incubation is where you fund bets on the future, deliberately walled off from the core business — new products, new markets, industries you think might matter in a couple of years. The whole point of keeping them separate is that they’re not on the hook for this quarter’s numbers.
The second you hold an incubation project to the core’s ROI bar, you kill it.
Your job here is to give these ideas room, resources, and air cover, and to be patient about it. Some won’t bloom for a year or two. That’s fine. You’re planting, not harvesting.
Transformation: the one big thing you drive
Transformation is the part that’s really yours. It’s the single big initiative you personally drive as the agent of change — the thing everyone in the company knows is critical to the future, even though it doesn’t move this quarter’s revenue or ship in the current release.
When you’re in the Transformation Zone, the message to the company is simple: this is going to happen, and I’m personally on it. It might be expanding into a new international market, rebuilding your go-to-market motion, or changing how you price. Whatever it is, the company’s future runs through it, and you’re the one carrying it.
I’ve run this play my whole career. At UserTesting, when I first joined, the transformation was the go-to-market team — we redefined territories, built new segments, revamped the packaging. Once that was in motion and had settled into the Performance and Productivity zones, I moved on. The next project was expanding the product to reach new buyers; it started as an incubation bet, gained traction, and got promoted at a sales kickoff with a “now is the time” message that told the whole company where the focus was headed. After that, a full transformation of our pricing model became the big one.
One at a time
Notice those happened in sequence, not all at once. That’s the discipline. You cannot run three transformations in parallel without generating exactly the chaos your cadence exists to prevent. Pick one. Drive it until it’s running on its own inside the core business, then turn to the next.
Incubation gives you a standing pipeline of candidates; the hard judgment is knowing when to pull one out of “someday” and make it the company’s main focus. That call — and the confidence to rally everyone behind it — is one of the most CEO things you do.
It’s tempting to picture the job as oversight: keep the trains running, watch the numbers. At a scaleup, that’s half of it at most. The more important half is seeing where the market is going before the org does, and dragging the company toward it.
If you want the full model, Geoffrey Moore’s writing on Zone Offense is worth your time — but no framework drives change on its own. It only works if the CEO decides to be the one creating the right problems.
Comments welcomed: