

This is the weakness question wearing a friendlier hat, and the phrasing matters. "Growth" implies a direction of travel — the interviewer isn't asking what you're bad at, they're asking what you're currently working on and how you know it's working. Answer it as a pure weakness question and you'll sound flat; answer it as a trajectory and you'll sound like someone worth investing in.
Pick something that is real, adjacent, and in progress.
Real means it would show up in your last performance review. Adjacent means it sits next to the job rather than inside it — public speaking for a backend engineer is adjacent; writing correct code is not. In progress means you can describe what you're doing about it this month, not in the abstract.
Note what this rules out. "I'm a perfectionist" and "I care too much" are not real. "I struggle with the main duty of this role" is not survivable. "I'd like to be better at strategy" with nothing after it is not in progress.
"Communicating technical work to non-technical stakeholders. I'm comfortable in a design review, but last year I presented a migration plan to our commercial team and lost the room in about ninety seconds — I opened with the architecture rather than with why anyone should care.
What I've been doing since is writing the summary before the detail: one paragraph on the business impact, then the technical plan underneath, and I ask a non-engineer to read the paragraph before it goes out. I've done that for the last four significant proposals and all four got approved without a follow-up meeting, which wasn't my track record before.
I'd still say it's my weakest area relative to the rest of my work — I have to consciously do it rather than it being instinct."
"Getting comfortable killing things that are working slightly. I'm good at starting channels and bad at shutting them down, because there's always some attribution you can point at to justify keeping one alive.
It came to a head when I realised we were maintaining six channels and two of them were consuming about a third of my team's time for something like 4% of pipeline. Since then I've put a quarterly review on the calendar with a rule I have to follow: anything under a threshold I set in advance either gets a specific fix with a deadline, or it goes.
I've done that twice now and cut one channel. I found the second one genuinely hard, so I wouldn't say I've fixed the instinct — I've just built a process that doesn't rely on it."
"Knowing when to ask for help. My instinct is to keep going, and early on I spent most of two days on a data pipeline issue that the person sitting next to me solved in about twenty minutes once I finally asked.
My rule now is a hard forty-five minutes: if I haven't made progress by then, I write up what I've tried and post it in the team channel. Writing it up sometimes solves it on its own, and when it doesn't, the question is much easier for someone to answer.
It's better than it was, though I notice I still stretch the forty-five minutes when I feel like I should already know the answer."
"Delegating work I'm faster at than the person I'm delegating to. As a new manager my default was to take the tricky thing back, and it was hurting the team twice — they weren't learning it and I wasn't doing my actual job.
The change I made is that I now assign the work along with an explicit budget: 'this will take you three times as long as me and that's the point, come back on Thursday.' Saying the cost out loud is what stopped me quietly reclaiming it.
Two of my four reports have picked up things I used to own. The other two haven't yet, and honestly that's because I've been less disciplined with them, so it's still live."