CS Interview Coach

By grade

Civil Service Grade 7 Interview Questions: What Panels Expect and How to Pass

Published · Last updated


At SEO the panel asks whether you made the right calls about the work. At G7 the panel asks whether you decided what the work should be.

That is the hardest calibration shift on the ladder, and it catches excellent candidates: ownership, personal drive and being across the detail, presented as your core evidence at G7, read as a grade below.

The format

A panel of two or three, at G7 often chaired by a Grade 6 or Deputy Director. The advert names the assessed elements, typically three to five behaviours, frequently a presentation or technical exercise, plus strengths questions. Interviews run 45 to 75 minutes.

Each element is scored 1 to 7 against the G6/G7 descriptors, pass mark usually 4, appointments in merit order. Mechanics in how Civil Service interviews are scored.

Two G7-specific notes. Situational questions appear more often than at lower grades, because panels want to test judgement on scenarios you may not have met. And probing is deliberately firmer, because holding positions under challenge is part of the job. Concede what is fair; hold what is not.

What "G7 level" actually means

The distinctive G7 evidence, across every behaviour:

  • Direction set or changed. You framed the problem, proposed the approach, or challenged the plan that no longer fit, rather than delivering a direction someone above you set.
  • Decisions on incomplete evidence. The unknown named, the assumption stated, a trigger set for revisiting, rather than analysis that made the answer obvious.
  • Senior influence. A Deputy Director's position moved, a no delivered upward with reasons that held, another department negotiated with, rather than stakeholders kept informed.
  • Delivering through leaders. Whole workstreams delegated to SEOs and HEOs with real authority, rather than personally driving everything important.
  • Wider context in your reasoning. Ministerial priorities, spending pressure and other departments' equities, visible in why you chose rather than recited as scene-setting.

The test for every example: whose strategy was it? If the honest answer is "my G6's, and I delivered it brilliantly", the example will keep collapsing to SEO however well you tell it.

The behaviours you're most likely to face, and the G7 trap in each

  • Leadership. The trap is team-management evidence: objectives, one-to-ones, conflict resolved inside the team. All real, all SEO. The G7-distinctive evidence is defending under pressure, with your team's call challenged from above and you standing behind it.
  • Making Effective Decisions. The trap is presenting neutral options for a board to choose. Excellent staff work, decision handed upward. G7 evidence forms a position, states what it did not know, and defends the call.
  • Seeing the Big Picture. The trap is aligning to strategy instead of shaping it. G7 evidence spots what the strategy has not noticed and changes it.
  • Delivering at Pace. The trap is re-planning your own workload heroically. At G7 pace is a system: priorities re-set across workstreams, resources moved between teams, pre-agreed triggers.
  • Communicating and Influencing. The trap is an audience too junior and stakes too low. G7 evidence briefs senior officials, negotiates across departments, and moves senior positions.

One question, two answers: hearing the gap

Question: "Tell me about a time senior stakeholders disagreed about the way forward on something you were responsible for."

FailSEO-level telling: caps around 3–4/7 at G7

"Mid-programme, we were asked to absorb a significant new piece of work, and we couldn't do both on the original plan. The two directors above the programme wanted different things — one wanted the existing delivery protected, the other prioritised the new work. I ran a planning exercise to establish what delivering both would require, and presented the board with three costed options: delay, phase, or additional resource. After discussion, the board decided to phase the new work. I re-planned the programme around that decision, communicated the changes to all teams, and we delivered on the revised timeline. The board thanked the team for managing a difficult period well."

PassG7-level telling: the same events, around 6/7

"Mid-programme, we were asked to absorb a significant new piece of work we couldn't deliver alongside the original plan, and the two directors above the programme wanted opposite things — one protecting existing delivery, one prioritising the new work, which carried a ministerial commitment. My first judgement was that this shouldn't go to the board as an open question: weeks of drift while 200 staff heard rumours. The evidence didn't settle it either — the volume forecasts for the new work were genuinely uncertain. So I formed a position: phase the new work behind a one-quarter slip, because the ministerial commitment fixed its start date but not its ramp-up speed — that flexibility was the only slack in the system — and I said openly that if volumes hit the top of the forecast range we'd revisit. Before the board, I went to each director separately: with one I traded a firm, written quarter rather than an open-ended slip; with the other I reframed phasing as protecting her commitment from launching into a half-finished operation. Both arrived backing the same recommendation, and it was agreed in twenty minutes. Six weeks in, volumes did hit the top of the range — the exact risk I'd flagged — and because the trigger was pre-agreed, I slowed the ramp-up under authority I already had and told both directors the same week, with the data. We delivered as agreed, and the pre-agreed-trigger approach became how the directorate handles contested resourcing calls."

Same programme. The first candidate does immaculate staff work and hands the decision to a committee. The second forms the position, moves both seniors before the meeting, states the uncertainty and pre-agrees what would change their mind, then handles the materialised risk at their own level. That is the entire SEO-to-G7 gap in one example.

The presentation, briefly

If your interview includes one, you are being scored on judgement and clarity for a senior audience, not coverage. Structure it as a submission: recommendation first, then the reasoning, then the risks and what you would do about them. Ruthlessly respect the time limit, because overrunning is itself evidence of the wrong kind. Expect the probing to focus on your recommendation's weakest assumption, which means you should know what it is before they ask.

Preparing efficiently: a five-step plan

  1. Pull the behaviours from the advert and prepare only those, plus strengths and any presentation.
  2. Choose five to seven examples where you set or changed direction. If you are an SEO applying upward, hunt for the moments you operated above your grade: the paper that changed the branch's direction, the position you argued a G6 into. Those are your G7 examples, even if they felt small.
  3. Run every example through the direction test. Whose strategy was it, what did you decide without complete information, which senior person moved, who did you deliver through.
  4. Pre-answer the G7 probes. "Why that option?", "What would have changed your mind?", "What was your role versus your team's?", "Knowing what you know now, what would you do differently?"
  5. Practise holding a position under challenge, out loud. G7 panels push back on purpose. Rehearse conceding what is fair and holding what is not, because the reflex has to be built before the room.

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 FreeStart a practice session

3 free practice questions · No card needed · Cancel anytime