Balkrishna Attal

Blog / Hiring & Careers

Hiring & CareersAugust 28, 2026·5 min read

The Interview Question That Tells Me Everything

After enough interviews on the hiring side of the table, one question predicts more about a candidate than the other forty-five minutes combined.

— views
"Tell me about a feature that didn't work." Look for what they did next, not what went wrong.

I've sat on the hiring side of enough product management interviews now that certain patterns stop feeling like coincidence. Candidates prepare hard for the questions they expect — "walk me through a feature you shipped," "how would you prioritize X," "design a product for Y." Most people have a polished, rehearsed answer ready for all three. The question that actually separates people isn't any of those. It's the follow-up nobody rehearses:

"Tell me about a feature you shipped that didn't work. What did you do next?"

Why this question specifically

Every candidate has shipped something that worked — or claims to have, and it's genuinely hard to tell from a success story alone whether that's judgment or luck. A failure story is much harder to fake convincingly, because it forces two things a polished success story can hide: whether the person actually owns the outcome, and what they consider "done."

Three answers, and what each one tells me

The blame-shifted answer

"Engineering scoped it wrong," or "leadership changed the requirements halfway through," or "marketing didn't promote it properly." Sometimes these things are even true. But when the entire explanation for a failure lives outside the candidate, that's the signal — not the specific excuse, but the shape of it. A PM's job is influence without authority; if nothing is ever their fault, they're probably not claiming much responsibility when things go right either.

The vague-lesson answer

"I learned the importance of talking to users earlier." True, probably. Also said by roughly every candidate I've interviewed, usually without a next sentence. This tells me the story is real but under-examined — they experienced the failure without doing much thinking about it afterward.

The answer I actually want

A specific, sometimes unflattering detail about what they missed, followed by a specific, concrete description of what they changed about how they work afterward — not "I learned to listen more," but "I now write down the three assumptions I'm least sure about before I greenlight anything, and I go find evidence for the shakiest one first." That's not a lesson. That's a system, built out of a real scar.

What I'm actually testing for

Not whether they failed — everyone has. I'm testing whether they define "done" by shipping or by outcome. Someone who defines their job as shipping features will describe the launch and stop there. Someone who defines it by outcome will keep talking past the launch, into what happened after, because that's genuinely where the story lives for them.

If you're preparing for this

Don't reverse-engineer a "good" failure story to have ready. That's exactly as transparent as it sounds from the other side of the table. Instead, actually do the thinking once, honestly, about a real miss — what you'd have needed to know earlier, and what specifically changed in how you work because of it. If you can't find that second part yet, that's worth sitting with before the interview, not during it.

This is the kind of thing a single, well-spent thirty minutes can sharpen a lot — figuring out which of your own stories actually has that second half, and which ones you're still telling as though the ending is "and then I learned a lesson," full stop.

Stuck on something like this yourself?

I take on a small number of 1:1 calls, career conversations, and resume reviews each month.

See how we could work together
BA

Balkrishna Attal

A product leader with a background in civil engineering. Writes about product decisions, careers, and moving into product from somewhere else.