Decide on total cost, not usage alone: support load, maintenance drag, and the surface it adds to every future change, weighed against the revenue and retention genuinely attached to it. Then sunset in stages. Stop new users getting it, contact affected accounts directly with a date and a migration path, deprecate the API with a long window, and remove only once usage is flat.
Why interviewers ask this
Killing things is a senior behavior, and most candidates have only ever added. Interviewers want evidence you have counted the ongoing cost of a feature, not just its usage, and that you know a sunset is a communications project more than an engineering one. Mentioning direct contact with affected accounts rather than a changelog post is the detail that separates people who have done it from people who have read about it.
How to structure your answer
- Weigh total cost of ownership against genuine attached value.
- Check whether the users would actually leave without it.
- Lay out the staged sunset in order.
- Emphasize direct contact with affected accounts over a changelog.
Example answer
I look at total cost rather than usage, because usage alone always makes removal feel too risky. So I add up the support tickets, the maintenance time, and the drag it puts on every other change, since an old feature makes every refactor near it more expensive. Against that I put the revenue and retention genuinely attached to it, which usually means checking whether the accounts using it would actually leave. Then the sunset is staged, and the stages matter more than the decision. Stop new users getting it, contact the remaining accounts directly with a date and a path to something else, deprecate the API with a long window and a warning in the response, and only remove when the graph is flat. The thing that goes wrong is announcing by changelog and hoping. Anyone with a real dependency deserves an email and a person they can talk to about it.
Walking into this interview soon? GhostPilot listens to your live call, spots the question the moment it is asked, and puts a structured answer on your screen in real time. Try it on your next mock, or grab a $29 Session Pass, no subscription, for the real thing.
See how it worksFollow-up questions to expect
- How long a deprecation window is reasonable for an API?
- What do you do if a single large account refuses to migrate?
- How do you measure whether killing it was the right call?
Related product manager questions
Your interviewer will ask their own version of this. Paste your actual job description into the free Question Predictor and get the 20 questions that role is most likely to ask, with what each one is really probing.
Predict my questions