Last year, my team spent three months building a feature I was genuinely excited about. The research was solid. The design was clean. Engineering executed well. We shipped it, watched the metrics for two weeks, and the answer was clear: users didn't care. Adoption was flat. Engagement was negligible. The feature worked exactly as designed. It just didn't matter.
The hardest meeting I led that quarter was the one where I said, "I think we should kill it."
Why letting go is so hard
Three traps make it difficult.
The sunk cost trap: we've invested three months of engineering time, design work, and political capital. Walking away feels like writing all of that off. The rational move is to evaluate the feature on its future potential, not its past investment. But rationality is hard when you've watched the team pour themselves into something.
The ego trap: this was my idea. I championed it. I wrote the brief, sold it to leadership, and put my credibility behind it. Admitting it didn't work feels personal in a way that other product decisions don't.
The momentum trap: the team is energized. They're proud of what they built. Pulling the plug feels like telling them their work didn't matter. Even when everyone intellectually agrees that the data says stop, the emotional cost of stopping is real.
What teams that innovate do differently
The best product teams I've worked with don't just tolerate killing ideas. They build it into the process.
They set kill criteria before they start building. Before a single line of code, they define what success looks like and what would make them stop. "If we don't see X adoption in four weeks, we revisit." That agreement, made when everyone is objective, gives the team permission to stop when the moment comes.
They celebrate what they learned, not just what they shipped. A PM I respect told me her team runs "learning reviews" alongside launch retrospectives. "We ask: what did we learn that we didn't know before? If the answer is meaningful, the project was valuable, even if the feature gets killed." That reframe changes the team's relationship with failure.
They treat a pivot as judgment, not defeat. Stopping something that isn't working is a sign of discipline. Continuing to invest in something the data says won't work? That's the actual failure.
The discipline of letting go
Innovation isn't just about having great ideas. It's about having the discipline to let go of the ones that aren't working, so you can invest in the ones that might.
The feature we killed last year freed up the team to work on something smaller and less ambitious that turned out to drive three times the engagement. We wouldn't have found it if we'd kept pouring resources into the original idea.
The best product teams don't just ship fast. They learn fast. And they let go fast.