← All guides

INTERVIEW PREP

Competency-based interview questions: how they're scored and how to answer

Find out what each competency question is scored on, so you can give the panel the evidence it needs.

Say you're asked, “Tell me about a time you worked with people outside your team.” You pick your best project, the launch you led last year, and explain the deadline, the hard technical problem and the result. It's a good story, and it can still score badly.

The question was testing how you work with other teams. Your answer was almost entirely about your own work, so the panel had nothing to write down under that heading.

That's one of the most common ways to lose marks in a competency-based interview. A public-sector hiring manager on Reddit remembers hearing it as a candidate: “we loved the example but it didn't answer the question so you lost points”.

The fix usually isn't a better story. It's knowing which quality each question is testing, and telling the part of your story that shows it.

What's different about a competency-based interview

The questions sound like any behavioral interview: “Tell me about a time…”. The difference is the marking. Before the interviews start, the employer chooses a short list of qualities, its competency framework, and writes each question to test one of them.

The format is most common in the UK, and not only in the public sector. KPMG's interview advice says its hiring managers assess candidates against its values, competencies, skills and experience. Deloitte's early-careers interview checks how well you match the capabilities the firm looks for.

Analyst roles use it too. The FCA's graduate telephone interview covers motivation, values and competencies, and a data analyst on Reddit was told their interview would be “competency-based but also centred around technical data analysis”. The Civil Service and the NHS use the same approach, under the names behaviours and values.

What the panel writes down

While you answer, the panel is taking notes, and each note goes under one of the competencies being tested. When you finish, they score each competency from those notes. That's why the launch story above can feel good in the room and still score low.

Most employers keep their score sheets private, but the Civil Service publishes its scale, and it shows how strict the bar can be. Each behaviour is scored from 1 to 7, and a Defra recruiter explains that you need at least 4 on every one to pass.

In the Crown Office's scoring guide, a 4 means “adequate positive evidence and any negative evidence would not cause concern”. So one weak competency can cost you the job, and so can one worrying moment in an otherwise good answer.

The panel can only score detail it hears. “I helped improve the process” gives them nothing to note. “I timed each step, found approvals took four days and proposed same-day sign-off for small orders” gives them several things.

Finding the competencies before the interview

Because the panel scores against a fixed list, most of your preparation is finding that list and choosing stories for it.

Start with the job ad and the person specification. Public-sector ads usually name the behaviours being assessed. Private employers tend to repeat the same few qualities through the job description.

Write the list down in the employer's own words, and use those words in your answers. A civil servant on Reddit says it shows you're paying attention to the job's needs.

Employers often use different names for the same quality. If a job ad asks for “working together” or “collaboration”, it wants teamwork stories, so the teamwork advice below applies.

Then choose two real stories for each competency, and make sure at least one of them is about something that went wrong. Deliveroo asks for examples of past successes and of what you learned when things didn't go well. If you're applying for a graduate scheme, stories from study and volunteering count.

Making one story fit different questions

You won't have a separate story for every competency, and you don't need one. A good project usually holds evidence for several. What changes between questions is which part you slow down on.

Take one project. A data analyst's team is about to launch a new pricing page, and the analyst finds that the test results behind the launch are skewed by a tracking bug.

Asked about teamwork, such as “Tell me about a time you worked with people outside your team”, the answer spends most of its time on the people:

“The fix needed engineering and the product manager, and neither had seen the data. I walked them through two charts showing where the numbers split. The engineer agreed to patch the tracking that day, and I asked the product manager to move the launch by a week. I sent a short update each day until the re-run finished.”

Asked about decision-making, such as “Tell me about a decision you made without all the information you wanted”, the same project is told through the choice:

“I had three options: launch on time and correct it later, delay a week and re-run the test, or launch to half of users. The bug affected about a fifth of sessions, so the result could easily flip. I recommended the delay, because a wrong price would be harder to undo than a late page.”

The facts don't change. The panel hears different evidence because you chose which part to tell in detail. Try this with your strongest project: tell it once for each competency it could prove, and notice which sentences you keep each time.

The six competencies panels ask about most

These six come up most often in UK employers' lists. Use them to check that each of your stories has the part that scores.

Teamwork

Teamwork questions, such as “Tell me about a time you worked as part of a team to achieve a goal”, look for what you added to a group. In tech and data teams that often means working across functions, which Monzo's data team calls “a really important element of success”.

A story told entirely in “we” leaves the panel nothing to mark, but the opposite hurts too: a reply on Reddit warns that too many “I”s can make you sound like you're not a team player. Say what others did, then what you did.

Leadership

Leadership questions, such as “Tell me about a time you motivated people who didn't report to you”, check whether people followed your direction. You don't need a title: leading a project, a shift or a student society counts.

What scores is how you won people over, including someone who wasn't on board. A list of your responsibilities gives the panel little to mark.

Communicating and influencing

These questions, such as “Tell me about a time you explained something complex to someone without your background”, test whether you changed what someone understood or decided.

Say who the audience was, how you adapted the message for them and whether they acted on it. “I communicated clearly” isn't evidence.

Problem-solving and decision-making

Questions such as “Tell me about a difficult decision you made” are about how you reasoned. The decision itself matters less.

Walk through the options, the evidence and the risks, as the pricing-page answer above does. Jumping straight from problem to solution shows little judgment.

Adaptability and change

Some of these questions ask how you coped with change, such as “Tell me about a time you had to adapt to a major change at work”. Others ask how you started one, such as “Describe a time you suggested a better way of doing something”.

Either way, the panel wants to hear what you did, not just that you put up with the change. Complaining about it, even mildly, counts against you.

Resilience and working under pressure

This group covers pressure, competing priorities and setbacks, such as “Describe a time you had to juggle competing priorities”. Panels want the choices you made, not how hard you worked.

Say what you prioritized and what you dropped or renegotiated. “I worked late every night” isn't a choice. For setbacks, see Tell me about a time you failed.

How the expected answer changes with seniority

Panels judge your story against the level of the job. For a junior role, a careful decision within your own work is enough. For a senior role, they expect decisions that affected other teams, and how you brought people with you.

Example answers

Each role below answers a different competency question, so you can see how both the question and the evidence change with the job. The answers are made up, but each stronger one includes the details a panel could score.

In each made-up stronger answer, you can point to the sentences that prove the competency named above it. Try the same with your own answer: if you can't point to them, the panel has nothing to score.

Software engineer

Competency: problem-solving and decision-making.

Question: “Tell me about a decision you made without all the information you wanted.”

Weaker

“During an outage our database was overloaded and we didn't know why. I decided to roll back the latest release, which fixed it. I'm comfortable making quick decisions under pressure, and it worked out well.”

The decision is there, but not the options, the risk or why a rollback was the right call. “It worked out” describes luck, not reasoning the panel can score.

Stronger

“Our checkout database hit full CPU on a Friday evening, and orders were failing. I was on call. We'd shipped a release that afternoon, but it hadn't changed anything in the database, so I wasn't sure it was the cause.

I had two options. A rollback took ten minutes and was low risk. Finding the cause first could take an hour while orders kept failing.

I rolled back, and told the incident channel I wasn't sure it would help. I said that if CPU hadn't dropped in fifteen minutes, I'd scale up the database instead.

It dropped within five. On Monday we found a query in the release that skipped an index. I added a missing-index check to our code review, and it has caught two queries since.”

The options, the risk of each, the fallback and the follow-up are all there, which covers most of what a decision-making panel looks for.

Follow-ups to expect: Why not scale the database first? Who else did you tell, and when?

Product manager

Competency: communicating and influencing.

Question: “Describe a situation where you persuaded someone to change their mind.”

Weaker

“Our sales lead wanted us to build a custom feature for one big client. I showed him the data that proved it was a bad idea, and he agreed we should stick to the roadmap. Data is the best way to win an argument.”

The sales lead's reasons are missing, and “proved” and “win” describe a contest, not influence. There's little for the panel to score on listening or how the message was adapted.

Stronger

“Our head of sales wanted a custom export for one large client whose renewal was due. Building it would have delayed a feature that about forty smaller customers had asked for.

Before arguing, I asked him what the client actually needed. It turned out their finance team needed one monthly report in a set format, not a general export.

I showed him two options on one page: the full export, which would delay the roadmap by six weeks, or a scheduled report we could set up in three days. I also offered to join the renewal call.

He chose the report. The client renewed and the roadmap feature shipped on time. Since then, he brings me client requests before promising anything.”

The other person's reasons are taken seriously, the message is shaped for him, and the result includes a change in how he works with the candidate.

Follow-ups to expect: What would you have done if he'd said no? How did you estimate the six weeks?

Product designer

Competency: adaptability and change.

Question: “Tell me about a time plans changed at short notice and you had to adapt.”

Weaker

“Halfway through a redesign, leadership changed the priorities and cut our timeline in half. It was frustrating, but I'm very adaptable, so I worked hard and we still delivered something good.”

It describes the change and a feeling, but no decision of the designer's. “Very adaptable” is a claim the panel can't score.

Stronger

“I was three weeks into redesigning our onboarding flow when the launch moved forward by a month to match a marketing campaign. The full redesign wouldn't fit.

I went back to our research and found most people dropped off on one screen, where we asked for company details. I proposed shipping a new version of that screen for the launch and leaving the rest for later.

I walked the engineers and the PM through the cut that afternoon, so nobody kept building toward the old plan.

We shipped on the new date. Onboarding completion went from 52% to 61%, and the rest of the redesign followed six weeks later.”

The candidate decides what to keep, uses evidence to choose and helps the team switch plans quickly, which is what adaptability indicators describe.

Follow-ups to expect: What did you leave out, and how did you explain that? What would you have done without the research?

Data analyst

Competency: communicating and influencing.

Question: “Tell me about a time you explained something complex to someone without your background.”

Weaker

“I presented a regression model to the marketing team. They didn't understand the statistics at first, so I explained it in more detail and answered all their questions. In the end they understood it.”

More detail is usually the opposite of what this audience needed. The panel can't tell what the analyst changed, or whether marketing acted on it.

Stronger

“Our marketing team wanted to know which channels brought in sign-ups. I'd built a model that gave each channel a share of the credit, but my first presentation, full of coefficients, lost them within minutes.

I asked the marketing lead what decision she had to make. It was how to split next quarter's budget.

So I rebuilt the presentation around one chart: roughly what a sign-up cost through each channel, with a range to show how sure we were. I kept the method in an appendix for anyone who wanted it.

She moved about a fifth of the paid social budget into search, and the cost per sign-up fell by around 15% the next quarter.”

The analyst finds out the decision behind the question, rebuilds the message around it, is honest about uncertainty and shows that the audience acted on it.

Follow-ups to expect: How did you check the model before you presented it? What would you have done if the range had been very wide?

Program manager

Competency: leadership.

Question: “Describe a time you led a group of people who didn't report to you.”

Weaker

“I led a cross-functional program with five teams. I set up weekly status meetings, kept a risk log and made sure everyone hit their milestones. The program launched on time thanks to strong leadership.”

These are routine program tasks, not a moment of leadership. Nothing shows how the candidate got people to commit when they couldn't tell them to.

Stronger

“I ran the move of our customer data to a new region, which needed work from five teams, none of them mine. Two months in, the payments team hadn't started, because the work wasn't in their goals for the quarter.

Instead of escalating straight away, I met their lead to understand why. Their quarter was full. Together we worked out which piece only her team could do, about two weeks of work, and I moved the rest to a platform team that had capacity.

Then I took the revised plan to both directors in one meeting, so the trade-off was agreed once.

Payments finished their part within the quarter. We completed the move a week later than the original date, and I flagged that delay to the program sponsor a month ahead.”

The candidate leads without authority: finds out why a team is stuck, reshapes the work and gets the decision made at the right level. The honest week's delay makes the story more believable.

Follow-ups to expect: Why not escalate first? How did the platform team feel about taking on the extra work?

Consultant

Competency: resilience and working under pressure.

Question: “Tell me about a time you worked under significant pressure.”

Weaker

“Consulting is always high pressure. On one project we had a big deadline, and I worked until 2am most nights for two weeks. I thrive under pressure, and the client was happy with the final deck.”

Long hours are the only action. There's no prioritizing and no conversation about scope, and “I thrive under pressure” is a claim, not evidence.

Stronger

“Two days before a steering committee, the client asked us to add a cost comparison for three more countries. I was the analyst on the workstream, and we had good data for only one of them.

I listed what we could do well by the deadline: one country in full, and two using published benchmarks, clearly labeled as estimates. I took that to my manager with an estimate of the hours, and she agreed it with the client that evening.

I split the benchmark work with a colleague so I could focus on the full country.

The committee had what it needed, the slide said plainly which figures were estimates, and we sent the full numbers the following week.”

The pressure is real, but the answer is about choices: what to cut, who to tell and how to keep the work honest. Those are the details a panel listens for in a resilience answer.

Follow-ups to expect: What would you have done if the client had insisted on all three in full? How did you decide what to hand to your colleague?

Values and strengths questions

Two other kinds of question often appear alongside competency ones in UK interviews.

Values questions ask how you've acted on the employer's values, which the FCA and the NHS both test. Answer with a real example, but give more weight to why you acted as you did.

Strengths questions, such as “What motivates you?”, ask what you enjoy and do well. Keep the answer short and natural, without STAR.

Practicing your answers

A story can look relevant in your notes and still drift away from the question when you say it, which is how the launch story at the top goes wrong. So practice out loud, and ask someone to interrupt with “What did you do next?” and “What would you do differently?”

Some UK employers also expect a reflection at the end: the Bank of England uses STARR, which adds it to STAR, so be ready to say what you learned. STAR method interview examples shows where STAR answers still go wrong.