Why shouldn't a CEO ask the CPO for an AI roadmap?

Because a dated list of AI features is a feature parade. It rewards the CPO for shipping, not for moving an outcome, and it always looks short to a board that's already rushing. Ask for the outcome each bet moves and what it costs per outcome instead.

Is hours saved a good AI measure for product?

It's the weakest common one. In the Wharton and GBK data, 40 percent of departments use hours saved and 27 percent use net-new revenue, yet departments tracking at least one growth measure reported strongly positive returns 35 percent of the time against 13 percent for efficiency alone. Product's version of hours saved is velocity and PRs merged.

How far behind are product teams on AI?

Further than most CPOs assume. Wharton and GBK found Product/Engineering uses agents at 62 percent, close to other departments, but only 16 percent have AI built into core processes, 8 percent have transformed workflows, and 19 percent report strongly positive ROI, tied with Legal for last.

Why does cost per outcome belong in the CEO conversation?

Because buyers are asking. Wharton found 64 percent of managers and above, and 74 percent of VPs and up, say rising token, license, and API costs are a major barrier to using AI more. A story without unit cost doesn't survive the CIO's questions.

What happens if the CPO doesn't own the AI story?

The CEO finds someone who will. Equilar reported median chief AI officer pay at $1.9 million in 2026, with 60 percent appointed in 2025 or 2026, even though fewer than 10 percent of CEOs and board members in BCG's survey thought a chief AI officer should lead AI strategy.

What Should a CEO Hold a CPO to on AI?

What should a CEO hold a CPO to on AI?

THE SHORT ANSWER

Four things, and none of them is a roadmap. A growth measure named before each AI bet ships, because hours saved, velocity, and adoption tell a CEO that product got faster, not that anything a customer cares about moved. Cost per outcome, reported next to the outcome. A straight answer on how product's own work has changed, since the October 2026 Wharton and GBK Collective study found only 16 percent of Product/Engineering departments have AI built into core processes, the lowest of any department. And one bet killed every quarter, with the reason. The CPO's side of the deal is one sentence per AI bet: the customer outcome, the evidence, the cost per outcome, and what was stopped to fund it.

The AI story belongs to the CEO, who's being graded on it: CEOs in BCG's 2026 survey estimated 35 percent of their performance review depends on AI ROI. The CPO's job is to give them a story that holds up, and the CEO's job is to refuse the two that don't, a roadmap of AI features and a count of hours saved.

What I'd hold a CPO to is short. A growth measure before each bet ships. Cost per outcome next to the outcome. An honest account of how product's own work changed, which matters because Wharton and GBK found product teams are where AI is least built into the work. And one kill a quarter.

The full argument is Your CEO Is Graded on AI. Hours Saved Won't Save Either of You., part 2 of The CPO Transition. Part 1 is Somebody Has to Steer: The CPO Now Owns the Operating Model, and the board version is The Investor and Board Narrative for AI-Era SaaS.

Last reviewed 2026-10-15 · 1 min read

What should a CEO hold a CPO to on AI?

SOURCES

THE LONG VERSION

RELATED ANSWERS