CTO · VP Engineering · Head of Engineering
AI Adoption Pressure: The Board Wants an Answer
The board member who DMs why are we not at 100% Copilot doesn't want a task force. They want a number, a policy, and a reason to stop worrying about it. Meanwhile your ICs are on Cursor and Claude already, with or without a policy, and your platform team is three weeks into comparing models.
Adoption percentage is the wrong number to hand them, and it's the one most CTOs reach for first.
What the board member is really asking for
Adoption percentage measures compliance. It rises when people open the tool and says nothing about whether the work got better, which means its only real future is embarrassing you: 90% adoption sitting next to a flat delivery chart while somebody asks the obvious question.
What survives that follow-up is a tiered answer, because at your scale AI isn't one decision. It's three, with different owners and different kinds of risk.
Individual productivity: opt-in, measured on cycle-time delta rather than seats filled. Team workflows: spec-driven and review-gated, because this is the tier where quality either holds or quietly stops holding. Product-embedded AI: one or two real bets, each with a kill criterion written before you start, since the expensive failure here is the pilot that nobody is willing to end.
The number I'd put in front of a board is rework and revert rate on AI-assisted work. Flat or falling while volume rises means you have leverage. Climbing means you have motion dressed as progress, and it's a great deal cheaper to learn that in Q2 than to explain it in Q4.
By the next board meeting: a written policy, one dashboard, and an answer to how much of this are we doing and why that holds when somebody pushes on it.
Working it 1:1
Bi-weekly, or weekly when there is a board meeting inside the month. I have sat in the seat: built Mews into a $2bn+ unicorn from 8 to 80 teams, led engineering at Manta before IBM acquired it, and spent 4+ years in the Bay Area as Principal Software Architect at Databricks. The pattern data comes from 3,400+ unique sessions with engineering leaders in 17+ countries. Rated 9.17/10 on average across 300+ reviews.
Pricing is public: 430 EUR per session, 361 EUR per session for ELC members, or 2,580 EUR per quarter (6 sessions over 3 months). The quarter drops to 2,166 EUR if you build the inquiry with my AI the inquiry: under 16 minutes, €361 a session on every package. You only pay for a session you rate 7 out of 10 or higher. Full pricing.
Related reads
The other situations CTOs bring
- The board wants a delivery number →
- First 90 days in an inherited org →
- Technical due diligence before a raise →
- Every decision routes through you →
- A sounding board without politics →
All six, plus the questions CTOs ask most: CTO coaching & mentoring