Dreambeans downloads dropped 84% in October, and Google took it to India anyway. The question that tests how you report a bad number.
Read the three framings side by side. Nobody in the article disputes the October figure. Sensor Tower states it, TechCrunch publishes it, and Google does not contest it. What differs is what each party puts next to it: the outlet puts a caveat on the comparison set, and the company puts a demand signal from a different market. The number is fixed. The frame is a choice, and that choice is where honesty either survives or quietly leaves the room. Every program manager eventually holds a status report where the true number looks bad and the question is not whether to say it but what to say beside it. That is why a product launch story is also, exactly, an interview question.
Why they ask it
The interviewer is not asking whether you have ever had a bad number. Everyone has. They are asking three narrower things. First, whether you will surface bad news early and on your own, before someone else surfaces it for you. Second, whether you know the difference between framing and spin: framing adds context that helps a decision-maker act, while spin swaps the comparison until the number stops looking like a problem. Third, whether you can hold the number and its interpretation together in one breath, so the audience gets both without having to pull either out of you. A manager who can do this is a manager whose reports can be trusted when things are going well, which is the only time trust actually pays.
The trap
There are two ways to fail this question, and they look like opposites. The first is the buried number. The candidate describes a status deck in which the honest figure appeared on a late slide, after several pages of context, wins, and roadmap. They call this "setting up the story." The interviewer hears a number that was hidden until the audience was tired. The second failure is the swapped denominator. The candidate explains that the metric looked bad against the original target but fine against a revised baseline, and they led with the revised one. This is the Dreambeans move in miniature: the comparison set changed, and the audience was not told that it changed. Even when the new comparison is more fair, as TechCrunch's caveat arguably is, presenting it without naming the swap turns a defensible frame into a concealment.
A quieter third failure is abdication. The candidate reports the raw figure with no interpretation and no recommendation, calls that transparency, and leaves the sponsor to work out what it means. A number without a frame is not honesty. It is a hand-off of the hardest part of your job.
Applying STAR-T
Situation. Set the stage in two or three sentences, and name the number early in your answer the same way you should have named it in the room. "Our internal rollout of a new deployment pipeline was meant to be at majority adoption by the end of the quarter. Six weeks in, adoption was well under a third of target, and two of the largest teams had not migrated at all." The interviewer should know the shape of the bad news before you have finished the situation.
Task. Say what the report had to achieve, which is never "make the number look better." It is usually to get a decision made. "My job was to give the steering group an accurate picture in time for them to decide whether to extend the timeline, add migration support, or narrow the scope to the teams that were already moving."
Action. This is where the framing lives, and the strong answer describes it mechanically. The number went first, on the first slide, against the original target rather than a revised one. Then the context: which teams were behind and why, and what was different about the teams that had migrated. Then the one comparison that actually mattered for the decision, with the comparison itself named out loud. "I showed adoption against the original target, and separately against the migration pace of the previous pipeline change, and I told them I was showing the second one because it suggested the lag was normal for the first six weeks rather than a signal to stop." Finally, a recommendation with a date attached.
Result. Report what the audience did, not how they felt. "The steering group extended the timeline by one release cycle, funded two engineers for hands-on migration help, and asked for the same chart at the next review. Adoption passed the original target one cycle late." If the number stayed bad, say that too. A result that includes a second bad month, honestly reported, is stronger than a tidy recovery nobody believes.
Trade-off. Name what the framing cost. Leading with the bad number against the original target made the program look worse than a kinder baseline would have, and it invited a scope-cut conversation you would rather not have had. Say why you accepted that cost: the alternative was a sponsor who found out about the gap from someone else, and a report that could not be trusted the next time it said things were fine.
The follow-up that breaks weak answers
"What was the number the following month?" A candidate who framed honestly has a ready answer, because the frame was a forecast they stood behind and then checked. A candidate who spun the number stalls, because the frame was built to survive one meeting, not the next one. The second question an experienced interviewer asks is quieter: "Who heard the number before your sponsor did?" If the answer is nobody, the report was early. If the answer is a long list, the sponsor was last to know, and the framing was damage control dressed as transparency.
Score your answer against the director’s bar
Q: Tell me about a program status you had to report where the honest number looked bad. How did you frame it?
Bank this story now, while you still remember the number you were tempted to leave out, and rehearse it before the follow-up finds you. Try it free →