There’s a moment most PMs recognize.

You walk into a roadmap review ready to talk product: the user pain, the validation, the demo, the edge cases.

Then the room flips.

A senior leader asks something that sounds simple, but kills the momentum:

• “How does this move the business?”

• “When do we see impact?”

• “What’s the tradeoff if we don’t do this?”

And suddenly you’re no longer discussing a product. You’re defending a bet.

I loved Elena Luneva take:

"For a long time, I thought this meant execs 'didn’t get product.' The real answer (...) I was speaking in product. They were listening in business."

In roadmap reviews, execs evaluate four things. If you can answer all four clearly, approval gets easier:

—

Q1: What is it?

They are asking: Can I repeat this back accurately in one line?

Most PMs are strong here. We can explain the workflow, show mocks, and demo prototypes. But what we really need to do it to define the expected business outcome.

Example answer: “This is Auto-Triage for support agents. It routes tickets by intent and risk, so urgent cases hit the right queue on first pass reducing OPEX by 1%."

—

Q2: Why does it matter?

They are asking: Which business lever moves, and what is the mechanism?

PMs often answer in product language - engagement, friction, user needs. Execs are listening for business outcomes - revenue, cost, margin, risk reduction.

Example answer: “This matters because it reduces cost by x% resulting in y revenue retention via faster triage, which shows up as lower cost per ticket within a quarter.

—

Q3: When do I get it?

They are asking: When do we learn, and when can we change course?

Roadmaps don’t fail because timelines slip. They fail because leaders stop believing them.

In AI products, quality is probabilistic. Drift is real. Integrations surprise you. So, give them a learning timeline and a decision point.

Example answer: “First signal in 2 weeks from a 5% pilot, credible read in 6 weeks after expansion, scale-or-stop by week 4. Here's how we’ll validate it (...)"

—

Q4: How much does it cost me?

They are asking: What are we spending, what are we not doing, and what burden are we creating?

It sounds like finance. But cost is not just money. It's engineering capacity, opportunity cost, cognitive load, AI infra costs, maintenance, etc.

You do not need perfect numbers. You do need to show you are not ignoring reality.

Example answer: “This costs 2 engineers for 6 weeks plus labeling, then ongoing model spend and weekly eval maintenance. Tradeoff: onboarding improvements slip by 4 weeks this quarter.

—

Translating between product, finance, and exec judgment is what separates: PM → Senior PM → Director → VP / CPO.

You don’t need the title to practice the skill. But you do need the skill before the title shows up.

That's one of most important skills to master in 2026.

—

P.S. Enjoy this?

A full post (for everyone) with special templates (subscribers only): productcompass.pm/p/pro…