Today a product owner brought me a strong innovation idea for the business.

He wanted to pipe call-centre logs through agents so we could understand customer sentiment and which activities make people unhappy. From that, send real-time notifications to the human agents taking the calls. And when someone rings up unhappy about one of the online platforms or the services we offer, use what we know from the conversation to raise a high-quality ticket in the right domain, instead of a thin handoff that makes the next team guess.

Proper idea. Three connected moves, not a vague AI wave. I could already see the work packages: log access, sentiment, the alert surface, ticket quality, domain routing, the operational change for people on the phones.

I liked it. I told him that.

I still said no.

Not because the idea was weak. Because of the weeks it would take, and because in that same stretch my team could finish several smaller innovations that would reach people sooner than one large bet still being scoped.

He walked me through the logs, the live alerts, and the ticket piece for complaint calls. I asked questions. He had answers. The customer pain was obvious. I could feel myself wanting to say yes just because the conversation felt good.

Then I said: I like where your head is. I don't want this to own the next block of our time. I'd rather we keep shipping smaller improvements we can finish. If this still looks like the right large bet later, we can come back with the trade-offs written down.

He took that seriously. I hope he also heard I wasn't brushing him off.

Bad ideas are easy. You argue quality and leave.

Good ideas pull at you. Sentiment from call-centre logs, live pings to human agents, and better tickets on complaint calls sounds like the story you'd rather tell at the end of the quarter. Saying no can feel like you're the person slowing things down.

I've watched large innovation work soak up months while everyone still "believes in the vision" and it's hard to point at what changed for a customer. What disappears first is usually quieter: an onboarding fix, an edge case that burns support time, a tool that deletes an hour of weekly toil. That quieter list is what I was holding next to his pitch.

When something needs a large slab of calendar, I ask what we won't ship in that same slab. Today the answer was concrete enough. Several smaller things my team could have finished. His idea would still have been in socialising and scoping: data access, how noisy the alerts get mid-call, ticket quality bars, which domain owns which complaint.

I declined for that reason. Not because I've gone cold on big bets.

He deserved a straight reason. My team deserved the runway. Both of those were true at the same time, which is why these conversations are awkward even when they go well.