Tell me about a time you led a project when you were not the manager.
answer
- the gap nobody owned
- how you got the mandate
- a plan peers could commit to
- keeping it moving weekly
- one number, and who holds it now
basics
~10 sTests whether you create direction where none exists. Name the gap nobody owned, say how you got peers to commit without any authority over them, and end with the work finished and still running.
how to answer
5 beats- the gap nobody owned, and why it matteredTwo or three sentences, no more — this is about fifteen percent of your airtime. Say what was going wrong, who it was hurting, and make it clear the work was unassigned rather than delegated to you.
- how you got the mandate without the titleName whose agreement you asked for and what constraint you accepted in return, such as not slipping your committed work. This one beat is what separates leading from freelancing, and most candidates skip it entirely.
- the plan peers could actually commit toDescribe how you scoped it small and what you asked of each person: specific, sized, and cheap enough to say yes to. Mention if you spoke to people individually before proposing it in a group setting.
- how you kept it moving week to weekThis beat plus the previous one should carry roughly sixty percent of your airtime. Cover how progress was made visible, what you did when someone slipped, and which unglamorous parts you took on yourself.
- the result, and what still runs todayGive one before-and-after number, then name the person, rota or automated check that carries it now. Finish with one sentence of reflection on what you would structure differently, and stop talking.
your answer
5 story prompts- Pick an initiative you volunteered for rather than one your manager assigned, ideally within the last two years.
- Write down whose agreement you needed to start, and what constraint you accepted in exchange.
- List each peer you needed and the one specific thing you asked each of them for.
- Find one before-and-after number for the outcome, not a description of effort.
- Name what still runs today because of it, or say honestly that it stopped when you did.
draft and rehearse your own answer in a learn session
go deeper
This probes ownership and influence: whether you create direction in the absence of a mandate or wait to be given one. Interviewers want evidence that you can align peers who do not report to you, hold commitments without a reporting line, and carry work past the interesting part to a finished state. A strong answer shows initiative, visible coordination mechanics, and a measurable outcome that outlived your attention.
I was a mid-level Android engineer at a mobile agency, on an account staffed with nine engineers across two client apps. We had a stretch where one release started crash-looping on low-memory handsets; we hotfixed it inside four days, and then everyone rolled straight back onto feature work. Making sure it did not repeat was everyone's job, which meant it was nobody's. I asked the account lead for twenty minutes and said I wanted to own that for the quarter, with one measurable target and no extra people. She agreed on the condition that my committed feature work did not slip. The root cause was weight. We shipped every asset inside the base download, 74.3 MB of it, and installs were failing partway through on low-storage devices. I wrote a two-page plan with three concrete cuts and walked it round each engineer individually before I ever raised it in a team meeting, so nobody heard it first in a room. I asked each person for one specific thing, sized in days. We ran it as a standing fifteen-minute slot on Thursdays. I kept one visible list of who had what and what was blocked, and I took the dull parts myself — auditing the asset catalogue, chasing design for source files. By the end of the quarter that download was 41.6 MB and install-stage failures on low-storage devices dropped to under a fifth of what we had seen. Two of the checks live in the build pipeline now, so it cannot drift back quietly.
The mandate conversation and the accepted constraint are what make this read as leadership rather than a side project. Walking the plan round individually, and naming the unglamorous work taken on personally, are the mechanics interviewers listen for. It would downlevel if the result were stated as effort or the pipeline checks were left out.
By then I was the senior engineer on the platform side of a mobile agency, working across six client accounts rather than sitting on one. The trigger was a bad fortnight: two different apps shipped releases that would not complete installation on entry-level handsets in one large market, and both traced back to bundle weight nobody had been watching. Nobody was going to appoint someone to fix that. The accounts had separate budgets and separate leads. So I wrote a one-page memo framing it as client risk rather than engineering hygiene and took it first to the two leads whose apps had broken, because they had the scars and would carry it for me. With them behind it I asked for a standing half hour every second week and one named engineer per account. I kept it deliberately narrow. A per-app download budget agreed with each lead rather than imposed, a pipeline check that fails a build when a release adds more than the agreed amount, and a short report each cycle. In the first session I said out loud what we would not do — no shared component library, no refactor spanning accounts — because the fastest way to kill this was to let it grow into a platform programme. Median install size across the six apps came down from 68.2 MB to 44.9 MB over two quarters, and per-release growth fell from around 3.4 MB to under 0.6 MB. I handed the report to a rota after the second quarter and have not chaired it since.
The senior signal is in the sequencing and the subtraction: allies with scars recruited first, the framing shifted to client risk, and the scope publicly bounded in the opening session. Naming what the effort would not do, then handing the mechanism to a rota, is what carries it. Vague sponsorship or no handover would downlevel it.
Task scope is fine and expected. Show that you led one slice — sequenced the work, pulled in one or two peers, kept people informed — and that you finished it rather than handed it back half done.
Project or feature scope with several peers involved. The mechanics are the signal: how you scoped it, what you asked each person for, how you tracked commitments, and what you shipped. One measurable outcome is expected.
Cross-team or production scope. Show that you shaped the direction yourself, said out loud what the effort would not cover, absorbed the unglamorous coordination work, and left an owner or a mechanism behind so it survived you.
Org scope and a repeatable mechanism, not a single successful push. Show how you built sponsorship, how you knew when to stop or hand off, and what you deliberately declined so the effort stayed narrow enough to land.
saying these in an interview costs you the question
- Describing a project you were formally assigned to lead rather than one you stepped into
- We-language throughout, with no sentence naming what you personally decided or did
- Leading by escalation — handing coordination to a manager instead of aligning peers yourself
- Ending on effort rather than outcome, with no result anyone could measure
- Taking full credit while naming nobody else's contribution to a multi-person effort
- The initiative quietly fizzles and the answer stops before saying what became of it
- How did you get people to prioritize your project over their own work?Answer with mechanics, not charisma. Say what you asked each person for, how small the ask was, whose interest it served, and who you cleared the time with. If someone's manager had to agree, say so — going through the person's lead is a strength, not an admission of weakness. Avoid answering that everyone simply saw the value.
- What happened when someone committed and then missed?Show that you noticed quickly and handled it directly rather than either absorbing the work silently or escalating on the first slip. Describe the conversation, what you re-planned, and whether you dropped scope or moved the date. A candid example where you did it clumsily and learned from it usually reads better than a story where nobody ever slipped.
- Who owns it now?This probes follow-through. Name the person, rota or automated check that carries it, and say how the handover happened. If it died after you left, say so plainly and say what you would build differently to make it outlive you — that answer scores better than a vague claim that it is still going strong.
- What would you do differently if you ran that again?Pick one real structural regret, not a humble-brag about caring too much. Good material: you started too broad, you skipped an early conversation with someone whose team you touched, you tracked activity instead of outcomes. Keep it to two sentences and say what you now do instead.
## What this prompt is really testing Interviewers use it to separate people who need a mandate from people who manufacture one. Almost every meaningful engineering improvement — a migration, a release process, a quality push — starts life unassigned. The question asks whether you have ever been the person who noticed, framed it, got peers moving, and stayed with it after the interesting part was over. ## Common wordings you should recognise as the same prompt - Tell me about a time you led without formal authority. - Describe a project where you were the de facto lead. - Tell me about a time you drove something end to end that was not yours to drive. - When have you had to get work done through people who did not report to you? All of these want the same story; only the emphasis moves. - The **authority-flavoured** wordings want more on how you got peers to commit; - the **end-to-end** wordings want more on follow-through and the finished state. ## Weak versus strong, on the same story - A **weak version** spends ninety seconds on background, then says the team came together and it went well. - A **strong version** front-loads the gap in two sentences, spends the bulk of the airtime on specific moves — who you talked to first and why, what you asked each person for, how you made progress visible, what you cut when it slipped — and closes with a number and an owner. The single most common structural failure on this prompt is a rich setup and a ten-second result. Budget roughly: - fifteen to twenty percent on situation and task, - sixty percent on your actions, - and the remaining twenty-odd percent on the result and one line of reflection. ## Evidence that lands Three kinds carry disproportionate weight. 1. First, **the mandate conversation**: naming the person whose agreement you sought, and the constraint you accepted in exchange, proves you led rather than freelanced. 2. Second, **the shape of your asks**: small, specific, sized commitments from named peers show you understood that people without a reporting line to you will only honour asks that are cheap and legible. 3. Third, **the artefact you left behind** — a check in the pipeline, a rota, a line in a release checklist. Anything that survives your attention is stronger proof of leadership than the outcome number itself. ## Where the bar moves by level - At a **junior** level nobody expects you to have coordinated an organisation; leading one slice of work and finishing it is a complete answer, and the credibility comes from the finishing. - In the **middle band** the interviewer is listening for coordination mechanics and a delivered outcome. - At **senior** the scope should reach beyond your immediate team and the answer should include a scoping decision made in public — what the effort explicitly would not do — plus the mechanism that kept the result from drifting back. - At the **top band** the story should be about building a way for this class of work to get owned, including sponsorship and the things you turned down to keep the effort landable. ## Two traps 1. The first is picking a project you were actually assigned; interviewers ask a clarifying question and the story collapses. 2. The second is telling a story where your leadership consisted of raising the problem repeatedly until a manager assigned it to someone. Escalation is a legitimate tool, but if it is the whole plot, the answer is about noticing, not leading.