Do You Know When to Sunset a Feature?

From the Jason Calacanis startup checklist.

Every feature has a cost — engineering time to build, maintenance to keep running, and complexity added to the product. When a feature isn't working, the question is: how long do you let it breathe before cutting it?

Watch out for the sunk cost fallacy — the tendency to keep investing in something because of what you've already spent, not because of its future value.

Uber Pool: A Case Study

Uber Pool launched in 2014 as a carpooling option: cheaper rides for passengers, better route efficiency for drivers, lower emissions. The idea was sound. The economics were not.

Pool was lower-margin than standard Uber rides and never turned a profit. Uber sunset the feature in 2020, citing pandemic safety concerns — but it's widely believed the pandemic simply accelerated a decision that was already coming.

It took six years. That's six years of maintaining a feature that was eroding margins rather than improving them.

Questions to ask before sunsetting

• Is the feature being used? By how many users, how often?
• Is it profitable — or at least contributing positively to unit economics?
• Is the low usage a distribution problem (users don't know it exists) or a value problem (users tried it and didn't come back)?
• What would it cost to improve it to the point where it works? Is that a worthwhile investment?
• What is the opportunity cost — what else could the team build with that time?

Sometimes sunsetting is the right call. Ship the decision cleanly, communicate it clearly to users, and move on.