CS Interview Coach

By grade

Changing and Improving HEO Example Answers (Civil Service)

Published


Every behaviour has a home grade for its classic failure, and Changing and Improving's is HEO. This is where the suggestion trap does its damage — "I noticed the problem, I suggested a fix to my manager, and she took it forward" — because at HEO the trap is perfectly camouflaged: suggesting improvements upward is good workplace behaviour, it's what conscientious EOs are praised for, and candidates walk into the interview proud of exactly the sentence that caps their score. The HEO descriptors don't reward noticing and suggesting. They reward working with others to identify areas for improvement and simplify processes, explaining change and supporting people to adapt, taking managed risks. Identification and implementation, owned end to end — with the manager appearing where managers belong at HEO: at sign-off, presented with something already built.

What HEO panels score on this behaviour

A problem quantified before it was fixed. How much time, how many errors, how many people — measured, even roughly. The baseline is what makes your ending mean anything.

Ownership through the boring middle. The idea is worth almost nothing; the marks live in the testing, the adjusting, the persuading, the embedding. If your story's verbs stop at "suggested", so does the score.

A managed risk with a control. The HEO descriptor says it directly. Name what could have gone wrong and the thing you built so it wouldn't — a pilot group, a rollback plan, a deliberately narrow first scope.

People supported through the change. Someone found the change harder than you did, and you did something specific about that. The resistor converted is the single most reliable mark in this behaviour.

Technology chosen, not name-dropped. Where a tool features, the mark is the deliberate choice and the measured efficiency — not the brand name.

Questions this appears as

  • "Tell me about an improvement you identified and implemented."
  • "Describe a time you changed a way of working that others were used to."
  • "How have you used technology to make a process more efficient?"
  • "Tell me about a change that met resistance."

Probes to expect: "What was your role once the idea was agreed?", "What could have gone wrong?", "How did you bring people with you?" — implementation, risk, adaptation. The three HEO marks, in order.

Fail vs pass at HEO: the same logbook

FailVersion 1 — the suggestion trap, scores around 3/7

"Our team managed a pool of shared equipment — laptops, mobile devices, hearing loops — that colleagues booked out for visits and events. The booking system was a paper logbook by the equipment cupboard, and it caused constant problems: double-bookings, kit going missing with no record of who had it, and time wasted walking over to check availability. I could see this was inefficient and that a simple digital solution would fix most of it, so I raised it with my manager and suggested we move the log to a shared online booking sheet. She agreed it was a good idea and asked me to look into options, so I compared a few approaches and recommended using our existing workspace tools rather than buying anything new. She approved it, the new system was set up, and the team moved over to it. It's worked much better — double-bookings have basically stopped, and it's much easier to see where equipment is. My manager highlighted it in our team meeting as a good example of continuous improvement."

Trace the ownership: I raised it and suggestedshe agreed and asked me to look into optionsshe approvedthe new system was set up (by whom? the passive voice knows) → the team moved over (how? who handled the colleague who liked the paper log?). The candidate owns the noticing and a comparison exercise; every decision belongs to the manager and the implementation belongs to nobody. No baseline number, no risk considered, no one resistant, and the closing evidence is praise. This is the suggestion trap in its natural habitat — a real improvement, genuinely initiated by the candidate, told in a way that scores a 3.

PassVersion 2 — the same logbook, around 6/7

"Our team booked shared equipment — laptops, devices, hearing loops — through a paper logbook by the cupboard, and I kept seeing the cost: I tracked it loosely for a fortnight and counted four double-bookings, two items untraceable for days, and I estimated the walk-and-check routine alone at a couple of hours a week across the team. Rather than raise a problem, I built the fix first. I set up a booking sheet in the workspace tools we already had — deliberately, because the alternative of requesting new software meant procurement, cost and a wait, for a problem our existing licence already solved; I'd rather have a working answer in a week than a perfect one in a quarter. The risk I took seriously was the transition itself: a half-moved team is worse than either system, because nobody trusts either record. So I designed the switch to be safe — a two-week parallel run where the paper log stayed live as the fallback, with the sheet as the master, and I checked them against each other every couple of days. The resistance was real and reasonable: our office coordinator had run the paper log for years, and the change read as a criticism of her system and a threat to her role in it. I went to her before launching anything — asked her to co-design the sheet's layout since she knew the booking patterns better than anyone, and made her its administrator, which turned the person with most to lose into its owner. She spotted two fields I'd have got wrong. I took the finished sheet, the fortnight's baseline numbers and the parallel-run plan to my manager for sign-off — a ten-minute yes. Two colleagues never engaged with the emails, so I walked them through it at the cupboard, at the moment of booking, which is where the habit actually lives. After the parallel fortnight showed zero discrepancies, we retired the paper log. Six months on: no double-bookings, nothing untraceable, and the coordinator — its administrator — extended the sheet herself to cover the pool car, which told me the change had genuinely transferred."

The annotation:

  1. A baseline, counted. Four double-bookings, two lost items, two hours a week — a fortnight of loose tracking that makes the ending measurable. Version 1's "much better" is an adjective; Version 2's "zero in six months" is evidence.
  2. Built first, signed off second. "Rather than raise a problem, I built the fix first" relocates the entire story's ownership. The manager still appears — at sign-off, with a ten-minute yes, exactly where HEO puts her.
  3. Technology chosen with a because. Existing tools over procurement, working answer in a week over perfect one in a quarter — the technology descriptor ticked by a reasoned choice, not a tool name.
  4. The managed risk is the transition itself. Naming the half-moved-team failure mode and designing the parallel run against it is the HEO risk descriptor performed — a control, not a hope.
  5. The resistor becomes the owner. The coordinator co-designs and administers the thing that threatened her — supporting individuals to adapt to change executed as a structural decision, not reassurance. Her extending the system to the pool car is the proof it worked.
  6. Adoption treated as work. The cupboard-side walkthroughs for the two non-engagers — change explained where the habit lives. Version 1's "the team moved over" is this paragraph, skipped.

Calibrating either side of HEO

At EO, suggesting improvements and helping colleagues adapt genuinely scores — Version 1, tightened, is close to an EO answer. At SEO, Version 2 needs wider blast radius: processes simplified across teams, impacts weighed across a diverse range of end users, and an environment where others feel safe to challenge — see the SEO hub. The full grade ladder is in the Changing and Improving pillar.

HEO-specific mistakes on this behaviour

  1. The trap sentence. "I suggested it to my manager, who took it forward." If this sentence is load-bearing in your answer, restructure the telling or change the example.
  2. Ownership by passive voice. "The new system was set up" — by whom? Passive constructions are where donated ownership hides. Audit your answer for them.
  3. No baseline. Improvement claims without a before-number score as opinion.
  4. A frictionless rollout. Everyone adapting instantly means the adaptation descriptor goes unticked. Find the person who found it hard.
  5. Risk-free change. If nothing could have gone wrong, the managed-risk descriptor can't score. Name the failure mode and your control.
  6. Praise as the result. End on what changed and what it measures at, not on what your manager said in the team meeting.

If Changing and Improving is on your advert, check your answer for the trap sentence before a panel does — the AI coach scores against the real HEO descriptors and flags exactly where ownership leaves your telling. Three sessions free; the HEO Complete Guide calibrates all nine behaviours to the grade.

Common questions

Practise before the real thing

Reading how panels score is step one. The candidates who pass are the ones who practised saying their answers out loud — and got scored feedback against the official Civil Service Behaviours framework before interview day.

Start Practising Free

3 free practice questions · No card needed · Cancel anytime