Sample interview answers
Tell me about a time you solved a problem: sample answers
At my last job, customers kept calling back about late deliveries. I checked our call notes and found most calls came after a delivery estimate email that was always a day too early. I showed my supervisor, and the estimate was corrected. Repeat calls about deliveries fell sharply within a month.
To adapt it, keep the problem to one sentence, spend most of your answer on how you found the cause, and end with a result.
STAR problem-solving answers by setting
IT support analyst
Staff in one meeting room kept losing video calls mid-meeting, and the complaints had been going on for weeks. Restarting the equipment worked for a day at most. I logged every drop with the time and the call software, and saw they all happened when more than about twenty people were in the room. That pointed to the Wi-Fi rather than the equipment. I tested it, confirmed the room's access point was overloaded, and asked our network team to add a second one and wire in the conference unit. We had no further drops in the next three months, and I added the checks to our troubleshooting guide.
Why it works: It shows evidence gathering before fixing, rather than guessing.
Fresher, final-year project
For my final-year project, my model's results suddenly got much worse two weeks before submission, and I didn't know why. Instead of tweaking settings at random, I went back through my changes one by one. I found that a cleaning step I'd added was deleting rows with blank values, and those were mostly from one group, which skewed the data. I filled the blanks properly, retrained the model, and recorded what had happened in my report. My results recovered, my examiner singled out that section as a sign of good method, and I now check my data at every step rather than only at the end.
Why it works: It turns a student setback into evidence of methodical thinking.
Call center agent (Philippines, US client)
I noticed I was getting the same complaint several times a shift: customers charged twice after changing their payment card. Each time, I'd raise a refund, but the problem kept coming back. I checked with three colleagues and they'd seen it too. I collected ten account examples and sent them to my team leader with the pattern: the duplicate charge always happened if the card was changed within two days of billing. She escalated it to the client, who fixed the billing system. Until the fix went live, I wrote a short script so every agent could explain the charge clearly and promise the refund date.
Why it works: It moves from fixing cases to fixing the cause, with the right escalation.
Restaurant shift manager (US)
Our delivery-app orders were going out late on Friday nights, and we were getting poor ratings. Staff blamed the drivers, but I timed every app order for two weekends and found the delay was in the kitchen: app orders printed on the same printer as dine-in tickets and got buried. I asked the owner if we could move app orders to their own printer at the packing station and give one cook the job of watching it at peak times. Average wait times dropped by several minutes, complaints about cold food stopped, and our ratings on the app recovered within a month.
Why it works: It shows measuring the problem rather than accepting the obvious explanation.
Data analyst
Our sales and finance teams were reporting different monthly revenue figures, and senior managers had stopped trusting either. I was asked to work out which was right. I compared both reports line by line for one month and found three differences: refunds, a currency conversion date and orders cancelled after month-end. Neither team was wrong; they'd defined revenue differently. I wrote a one-page definition, agreed it with both heads of department, and rebuilt the report from one data source, with the definition printed at the top so nobody had to guess. From then on, there was one revenue figure in every meeting.
Why it works: It shows the problem was a definition, not a mistake, which is a mature insight.
Healthcare assistant (UK hospital)
An older patient on our ward was barely eating, and her meals were coming back untouched. I sat with her at lunchtime and found she couldn't open the packaging or cut her food with her arthritis, and she didn't want to ask. I told the nurse in charge, who referred her to the occupational therapist, and in the meantime I opened and cut up her meals and made sure she had a beaker she could hold. She started finishing most meals, and her food chart improved. I now check whether patients can manage their packaging, not just whether they ate.
Why it works: It shows observation and escalation within the HCA role.
School office administrator
As the school office administrator, I noticed parents kept missing important messages because they ignored the school's long weekly newsletter. Trips were going unpaid and forms weren't returned. I asked a few parents at the gate what they read and they all said short texts. I suggested we keep the newsletter but send a short text for anything needing action, with a direct link and a deadline. The head agreed to try it for one term, and I set up the texts using our existing messaging system. Forms came back much faster, trip payments were on time, and the office stopped spending whole afternoons phoning parents to chase replies.
Why it works: It shows asking the users rather than assuming the solution.
Retail supervisor (stock discrepancy)
Our store kept showing stock on the system that wasn't on the shelf, so customers ordered online for collection and we couldn't find their items. I looked at which products were affected and saw they were mostly returns. Returned items were being added back into stock at the till before anyone checked them, and damaged items were thrown away later without being removed from the system. I set up a returns bay, checked each item before it went back into stock, and made sure write-offs were logged. I explained the new process to every till colleague at our morning briefings. Failed collection orders fell noticeably within six weeks.
Why it works: It traces a customer-facing problem back to a process cause.
What makes these answers work
Problem-solving questions test how you think: whether you look for the cause before jumping to a fix, use evidence, involve the right people and check the result. A modest problem solved well scores better than a dramatic one told vaguely.
Each story follows STAR, with most of the words on the action, especially the step where the candidate found the real cause. Each also ends with a result that lasted, not just a one-off fix.
A fill-in template
Aim for about ninety seconds. Spend half of it on how you worked out the cause.
- Situation: At [where], [problem] was causing [effect].
- Task: I needed to [what you were responsible for].
- Action: I [how you investigated], found [the cause], and [the fix, including who you involved].
- Result: As a result, [measurable or visible outcome].
Mistakes to avoid
- Jumping to the fix without explaining how you found the cause.
- Choosing a problem you created and then fixed, without saying so.
- Ending without a result, or a result you can't describe.
- Making it a team story with no clear sense of what you did.
Frequently asked questions
What kind of problem should I choose?
A real work, study or volunteering problem where you found the cause and fixed it. It needn't be dramatic.
Can I use a problem I caused?
You can, but then it is closer to a mistake question. Be upfront about your part and focus on how you fixed it and what changed.
How long should my answer be?
About ninety seconds to two minutes, with most of it on what you did.