Tell me about a time you knowingly took a shortcut to ship something faster.
answer
- name the pressure and the stakes
- the two options, priced in hours
- who you told, in what words
- where the IOU was written down
- how the payback actually closed
basics
~20 sTests whether your shortcuts are deliberate and visible rather than quiet corner-cutting. Answer with one trade you chose on purpose, the alternative you priced, who you told, where the IOU was written down, and how it was paid back.
how to answer
6 beats- the pressure and what was actually at stakeTwo or three sentences on the deadline, the release or the incident, and what would have been lost by going slower. Keep this and the next beat to roughly a fifth of your airtime — a long setup is the commonest way this answer runs out of room before the payback.
- the two options and what each would have costSay the proper version out loud and price it in hours or days, then price the shortcut. This single sentence is what turns a corner cut into a decision, and it is the detail most candidates leave out.
- the call you made and who you toldState the choice in the first person and name who heard it before it shipped — a lead, a reviewer, a client contact. This beat and the next carry the bulk of your answer, around sixty percent of the time you spend.
- the IOU: where it was written and what made it duePoint at the artefact — a ticket, a comment, a line in a release note — and give the trigger rather than a vague someday. A condition that fires on its own beats a date that slips quietly.
- the payback and the resultSay when it was closed and what it cost to close, with one number if you have one. Give this and the reflection about a quarter of your airtime; an answer that stops at the ship is the version interviewers hear all day.
- what you would do differentlyOne honest sentence about the part you got wrong — an audience you skipped, a related gap you never checked, a cost you never quoted. Reflection is where the growth signal lives, so make it specific rather than modest.
your answer
5 story prompts- Pick a shortcut YOU chose rather than one you inherited, ideally within the last two years.
- Write down what the proper version would have cost in hours — you will be asked.
- Find the artefact that recorded the debt: a ticket, a code comment, a release note.
- Check your story has an ending — the day or trigger on which the debt actually closed.
- This can be the same story as your missed-deadline answer, re-angled onto the trade you chose.
draft and rehearse your own answer in a learn session
go deeper
Probes judgement and ownership under time pressure. Interviewers want to know whether your shortcuts are conscious decisions with a named cost, an audience and a payback plan, or quiet corner-cutting that only surfaces when something breaks. A strong answer shows the alternative you priced, the person you told, the place the debt was recorded, and how it was closed.
I was the junior tester on an agency team of nine delivering a client's checkout. One afternoon their payment provider started returning a response code we did not handle, and card payments were failing for a slice of customers — sixty-eight minutes of it by the incident timeline before we spotted the pattern in the error logs. The lead wrote a small fix for the unhandled code. The regression test behind it was my job, and it needed a provider stub I did not have yet. I said on the call that there were two versions: hold the release about forty-five minutes while I built the stub, or ship the fix now with nothing covering it and I write the test first thing. We shipped. Before I logged off I did two things. I opened a ticket named for the exact missing case and linked it to the incident, and I put it at the top of the next morning's standup instead of letting it drift into the backlog. I wrote the test the next morning; it took about half an hour with a fresh head, and it moved our payment-path coverage from 71 to 74 percent. What I would change: I never checked whether the same unhandled-code gap existed on the refund side. It did, and somebody else found it a week later.
The signal is that the shortcut was named aloud with its alternative and a price attached, not taken quietly. The IOU has a ticket, a link and a standup slot — a place it could not be silently dropped. It would downlevel if the payback were only an intention rather than a closed loop with a number.
On a ten-month engagement I owned the payment integration for a client's booking product. The contract was fixed-fee, and midway through we found their acquirer needed a mapping between our internal decline reasons and theirs — thirty-eight codes. The clean version was config-driven and loadable per acquirer, roughly six days. The shortcut was a hard-coded map in one file, a day and a half. I took the shortcut and made three things true first. The trigger went into the ticket in words: do this properly the moment a second acquirer is in scope. The same sentence went into the weekly client note, so their tech lead had it in writing and it was not an engineering secret. And I pointed the tests at behaviour — decline reason in, customer message out — so the eventual rewrite would not need a new test suite. Payment-path coverage held at 83 percent across both versions. A second acquirer landed about seven weeks later. The rewrite took four days because the tests already existed, and nobody argued about whether it was worth doing, because the trigger had been agreed while everyone was calm. The part I got wrong: I never gave the client a cost for the debt at the time, so it read as free until I turned up asking for four days.
The conditional trigger earns it, together with the trade reaching a non-engineer in writing — debt the payer has seen is a scheduled cost rather than a rumour. Pointing tests at behaviour shows the shortcut was chosen so it stayed cheap to reverse. The unquoted cost is a real, specific reflection rather than a modest one.
Own-task scope is fine: a test you skipped, a manual step you left manual. What has to be there is that you said the trade out loud to someone more senior and left a written trace, rather than deciding it silently at your desk.
Price both options in hours and pick with a reason a reviewer could check. Name the trigger condition that made the debt due rather than a vague someday, and show the trade reaching a non-engineer in writing.
Reach production risk and the team: what the shortcut could have broken for real users, what guardrail contained the blast radius while it stood, and what changed in how you release so the next one is not down to luck.
Talk about how deliberate debt gets decided and repaid across teams — a register with triggers and owners, an agreed tier of paths that never take shortcuts, and how the cost is made visible to whoever funds the work.
saying these in an interview costs you the question
- Framing every shortcut as something forced on you by someone else's deadline
- No record of the debt beyond a mental note
- The story ends at shipping, with no evidence the shortcut was ever paid back
- Presenting a shortcut nobody else knew about as pragmatism
- Picking a trade so trivial it carried no real risk to anyone
- Blaming the mess on a previous team or the client instead of owning the call
- Who else knew you were taking that shortcut at the time?Name the specific people and the channel, and be honest if the answer is nobody. Interviewers are testing visibility, not consensus: a shortcut a tech lead approved in a thread is judgement, the same shortcut discovered later in a review is a surprise. If nobody knew, say so and say what you would tell now.
- What would it have cost to do it properly?Give a number in hours or days, even a rough one, and say how you arrived at it. The point is that you actually priced the alternative before choosing. If you never estimated it, admit that — then say what you would compare against today, because a trade you cannot price is a trade you did not really make.
- Did the debt ever actually get paid back?Answer plainly, including when the answer is no. A closed loop is the strongest ending: the trigger fired, the work was scheduled, it shipped. If it is still open, say who owns it now and what keeps it visible. Pretending every shortcut you have taken was repaid reads as rehearsed rather than experienced.
- When would you make the opposite call?Show the boundary. Name the properties that flip you to doing it right the first time — irreversible effects, money or data at risk, a customer absorbing the failure instead of a colleague — and connect them to the story you just told, so the rule sounds derived from practice rather than recited.
### The prompt family This one arrives in several wordings and they all want the same evidence: *tell me about a time you took a shortcut*, *when have you traded quality for speed*, *give me an example of technical debt you took on deliberately*, *tell me about a time you shipped something you were not proud of*. The last wording is the trap variant — it invites confession, and candidates supply guilt instead of judgement. Answer all of them with the same story, angled at the decision rather than the regret. ### What the interviewer is sorting for Engineers fall into three groups, and the question separates them cheaply. The first cuts corners without noticing, and their story has no alternative in it — there was one way to do it and they did it. The second refuses to cut corners and describes a purity that has never met a delivery date; they usually answer with *I do not take shortcuts*, which lands as inexperience. The third group treats a shortcut as a priced decision with a creditor. That third answer sounds like: here was the pressure, here were two options and what each cost, here is the one I picked and why, here is who I told, here is where the IOU lived, here is the day it was closed. ### The evidence that carries the signal Four artefacts do most of the work, and each is checkable: - **A price on the alternative.** Not *it would have taken longer* — a number of hours or days, and how you got it. - **A named audience.** Somebody outside your own head knew: a lead, a reviewer, a client contact, a note in the release channel. - **A written record with a trigger.** The strongest version is not a dated ticket but a conditional one: *this becomes urgent the moment a second provider is added*. Dates on debt tickets slip silently; conditions fire on their own. - **An ending.** The debt was closed, or it is still open with an owner and you can say why. ### Weak versus strong, on the same story Weak: *We were under pressure so I hard-coded it, and yeah, it is still like that. It works fine.* Everything is passive, nothing was priced, nobody was told, and the ending is a shrug. Strong: *The clean version was about six days, the shortcut a day and a half. I took the shortcut, wrote the trigger into the ticket, put the same sentence in the weekly client note so it was not an engineering secret, and pointed the tests at behaviour so the rewrite later would not need new ones. The trigger fired about seven weeks on and the rewrite took four days.* Same decision; one of them is a professional making a trade, the other is a professional making an excuse. ### How the bar moves with level At the earliest level nobody expects you to own the call — they expect you to surface it. Skipping a regression test and saying so in standup is a complete answer. In the middle band the trade should be yours, priced, and communicated beyond the team. At senior the interesting part stops being the shortcut and becomes the containment: what stopped a bad shortcut from becoming an incident, and what you changed in the release path afterwards. At principal scope it is policy — which paths are allowed to take debt at all, how the register is kept honest, and how the bill reaches the people who fund the work. ### Two things to keep out Do not pick a shortcut with no downside; a trade with nothing at stake proves nothing. And do not let the story end at the ship. The payback is the half of the answer most candidates skip, and it is the half the question was actually about.