Grappling with Traction as a Startup CTO
I just finished Traction by Gino Wickman. It’s pitched as an “operating system” for entrepreneurs, and as much as I usually roll my eyes at business-book frameworks, this one hit close to home. Here’s the truth: being a startup CTO isn’t just building product and shipping code. It’s herding people, aligning vision, and keeping the wheels from falling off while you’re driving 120 km/h.
And that’s where EOS (Entrepreneurial Operating System) stung a bit. It’s basically a mirror showing you all the places you’re winging it and pretending it’s fine.
The Six Components (Through My CTO Lens)
Wickman splits a company into six components. On their own, none of them are revolutionary. But together, they create a forcing function: either you confront reality, or you keep living in chaos.
| Component | What It Means | CTO Lens (My Reality Check) |
|---|---|---|
| Vision | Clear direction for everyone | Stop moving the goalposts every sprint |
| People | Right people, right seats | Talent ≠ fit; wrong role drags everyone |
| Data | Run on key metrics | Scorecards > gut feels |
| Issues | Surface and solve problems | Fires repeat? That’s a system flaw |
| Process | Document & simplify workflows | Hate it, but chaos doesn’t scale |
| Traction | Rocks + meeting rhythm | Execution > big ideas |
Where EOS Smacked Me in the Face
Let me break down where this actually landed for me:
- Vision. I know where I want us to go, but I haven’t always made it crisp for the team. Engineers don’t need a new whiteboard sketch every two weeks. They need stable goals that don’t shift with the wind.
- People. “Right people, right seats” sounds like corporate fluff until you see how much drag comes from misaligned roles. As a technical founder, I’ve kept “almost fits” too long.
- Data. I live and breathe instrumentation. But EOS pushes it further: a company scorecard, not just technical dashboards. Five to fifteen measurables, reviewed weekly. If you don’t have the numbers, you’re not actually managing.
- Issues. I love solving technical fires. But EOS forces you to ask: why does this fire keep happening? That’s a discipline we skip way too often.
- Process. I hate documentation for documentation’s sake. But Wickman’s right: scale without process is chaos. The trick is documenting the 20% that matters, not building an encyclopedia.
- Traction. Rocks (90-day goals) and L10 meetings are the system’s metronome. Without rhythm, a startup drifts. When you drift, you die quietly while everyone’s “busy.”
The People Analyzer (Get It, Want It, Capacity)
EOS has a simple but brutal tool for evaluating fit: does someone get it, want it, and have the capacity to do it? Miss two out of three? It’s time to make a hard call.
Here’s how it flows:
flowchart TD;
A[Person] --> B{Get It?};
B -- No --> N1[Not a Fit];
B -- Yes --> C{Want It?};
C -- No --> N2[Not a Fit];
C -- Yes --> D{Capacity?};
D -- No --> N3[Not a Fit];
D -- Yes --> F[Right Seat];I’ve already run a few names in my head through this filter. It’s uncomfortable, but ignoring the misfits is worse.
Scorecard: The Forcing Function
EOS insists on a weekly scorecard: 5-15 metrics that objectively show whether the business is on or off track. No “feels.” No arguing. Just numbers.
Here’s what a draft scorecard for us might look like:
| Metric | Target | Owner |
|---|---|---|
| Onboarding completed | 90% | CS Lead |
| Churn signals flagged | < 5/wk | Ops |
| API error rate | < 0.2% | Eng |
| Shipments processed | +15% | Ops |
| NPS | > 45 | Product |
Notice how it blends product, engineering, and customer success. This isn’t just “is the backend on fire?” It’s: is the business moving?
EOS Flow at 10,000ft
The way Wickman stitches it all together is neat: Vision sets direction, People run with it, Data shows reality, Issues get solved, Process makes it repeatable, and Traction drives execution.
flowchart LR
V[Vision] --> P[People] --> D[Data] --> I[Issues] --> PR[Process] --> T[Traction]Simple. But try skipping one piece and watch the whole system wobble.
Where I Push Back
Not everything in EOS is plug-and-play.
- Process overkill. Documenting every workflow is a recipe for bureaucracy. Keep it lean, or it will suffocate us.
- Discipline is assumed. EOS only works if accountability is already in the DNA. Without it, the framework collapses.
- Rocks ≠ chores. Rocks have to tie directly to the vision, or they just become another to-do list.
I am not convinced EOS magically “fixes” culture. But I am convinced chaos doesn’t scale.
My Next Steps as CTO
Here’s how I’ll start threading EOS into our company without drinking the full Kool-Aid:
- Weekly scorecard. Seven to ten metrics max, reviewed Monday mornings.
- Quarterly rocks. Three priorities, not ten. If everything’s a priority, nothing is.
- People Analyzer. Run the leadership team through the GWC lens. Better to face it now than six months later.
- L10 meetings. Kill the status fluff. Meetings exist to solve issues, not read bullet points.
The Bigger Picture
EOS isn’t magic. It’s scaffolding. But as CTO, I can’t just make the product work. I’ve got to make the company work. That means embracing systems, rhythms, and accountability that feel boring compared to hacking out a new feature, but are the only way we grow past “scrappy startup.”
Make it work, make it work well, then make it pretty. EOS is one way to force “work well” into the operating rhythm. And at the end of the day, data over opinions.