CS Interview Coach
Start Free

By grade

Working Together Grade 7 Example Answers (Civil Service)

Published

At Grade 7 the Working Together descriptors contain a phrase that quietly reorganises the whole behaviour: actively build and maintain a network of colleagues and contacts to achieve progress on shared objectives. To achieve progress — the network is instrumental. This isn't collaboration as relationship quality, which carried the marks at HEO and SEO; it's collaboration as capability: relationships built before they're needed, maintained through periods where they deliver nothing, and then deployed — the right call to the right person unblocking the thing six months of formal process couldn't. Paired with it is the behaviour's most grown-up descriptor: challenge assumptions while being willing to compromise if beneficial to progress. Both halves, in tension, resolved by judgement — knowing which hills matter and paying for progress with the ones that don't. G7 panels score exactly that judgement.

What G7 panels score on this behaviour

A network with history and reach. Contacts across departments, professions or organisations, built and maintained over time — including before and between the moments they were useful. "I set up a stakeholder group for the project" is SEO; a network is older than any one project.

The network, deployed. A specific moment where a maintained relationship achieved progress formal routes couldn't — with the mechanism visible: why that person, what the relationship history made possible, what moved.

An assumption challenged at a senior table. Something everyone in the room treated as fixed, contested by you with evidence — the descriptor's first half, which requires the discomfort to be real.

A compromise chosen, with the reasoning. Ground given deliberately — not lost — because progress was worth more than the point. The mark is the explicit judgement: what you traded, what it bought, why that was right.

An environment where challenge flows back. The inclusive-environment descriptor at G7 scale: structures or behaviours that made it safe for others — including juniors and external partners — to contest your assumptions.

Questions this appears as

  • "Tell me about building relationships across organisations to achieve something."
  • "Describe a time you challenged the consensus in a senior group."
  • "Tell me about a compromise you made to keep something moving."
  • "How have you used your professional network to deliver an objective?"

Probes to expect: "How did you know that person?", "What did the compromise cost, and who paid it?", "What would have happened without the relationship?"

Fail vs pass at G7: the same standards group

FailVersion 1 — chairing well, scores around 4/7 at G7

"I chaired a cross-government working group set up to agree common data definitions for how our departments measured and reported on a shared policy area — six departments, each with established definitions embedded in their own systems and reporting. It was a challenging group to lead: every department had reasons to prefer its own approach, and previous attempts at alignment had stalled. I ran the group in a structured and inclusive way — clear terms of reference, rotating agenda ownership so every department's issues got airtime, papers circulated well in advance, and disagreements worked through methodically in the meetings. I made sure quieter members were heard and that larger departments didn't dominate. Over fourteen months we reached agreement on a common core of definitions, which all six departments adopted in the following reporting year. The group was held up as an example of effective cross-government working, and I was asked to advise two other groups on how we'd approached it."

Read it at G7 calibration and the gap is structural: this is an answer about process quality — good terms of reference, inclusive chairing, methodical papers. All real, all creditable, and the behaviour's G7 machinery is absent: no network exists (the group is the only relationship structure, created for the task), nothing is deployed, no assumption is challenged (disagreements are "worked through methodically", which describes facilitation, not challenge), and no compromise is owned — agreement simply emerges from fourteen months of good meetings. The candidate chaired well. Chairing well is what secretariats enable and SEOs do. A 4.

PassVersion 2 — the same group, around 6/7

"I chaired a cross-government group tasked with agreeing common data definitions across six departments in a shared policy area — the third attempt in five years; the previous two had produced polite minutes and no change. Before accepting the chair, I did the thing the previous chairs hadn't been positioned to do: I audited my own network against the problem. Over the preceding years I'd built and kept relationships across four of the six departments — an analyst counterpart I'd worked a data-sharing agreement with, a policy lead I'd sat on a professional network steering group with, others maintained through nothing more strategic than turning up to cross-government profession events and being useful when asked. Those relationships changed what the group could be, because they gave me what the formal structure couldn't: honest intelligence about why alignment kept failing. Three private conversations before the first meeting told me the real blocker was never definitional — it was that the largest department's definitions were hard-wired into a system mid-replacement, and any agreement requiring them to change before their new system landed was dead on arrival, which they'd never say in a formal meeting. That intelligence shaped my first act as chair: I challenged the group's founding assumption at the first meeting — the assumption, written into our terms of reference, that success meant all six departments converging on one definition set on one timeline. I argued, with the previous two attempts as evidence, that this assumption was precisely what had killed alignment twice: it made the largest department a silent blocker and everyone else hostages to their system timeline. The pushback was immediate — two departments' representatives had mandates built on full convergence, and the sponsoring director's office had announced 'a single set of definitions' publicly. This is where I chose my compromise deliberately: I gave ground on the timeline and the totality — a common core of eight definitions now, adopted by five departments immediately and the sixth at their system cutover eighteen months later, with a published mapping in the interim — and I held ground absolutely on the core's content, where I had evidence the definitions materially changed policy conclusions. My reasoning, which I gave the group openly: a complete agreement in eighteen months was worth less than 80% of one now, because the reporting year ahead would otherwise produce a third round of incomparable data — progress was the hill; completeness wasn't. Getting the sponsoring director to accept a revised public position took a deployment of the network again: the policy lead I knew in her department walked me through how to frame the phasing so it read as delivery rather than retreat, and reviewed my note to her before it went up. The revised mandate was agreed. The core definitions went live for five departments that reporting year, the sixth cut over on schedule, and — the part I'd argue matters most at this grade — the group didn't disband: it became a standing network of its own, now maintained by the members, which has since resolved two definitional disputes I only heard about afterwards. I also built the challenge norm in structurally: every meeting ended with ten minutes where any member could contest any 'settled' item without papers or notice — which is where two of our eight core definitions got materially improved, both times by the most junior people at the table."

The annotation:

  1. The network precedes the task. Relationships built over years, maintained through profession events and reciprocity — audited against the problem before the chair is accepted. This is the descriptor's actively build and maintain with the timeline that proves it.
  2. Deployment with visible mechanism. Three private conversations produce the intelligence — the real blocker — that no formal structure could surface. Why the network mattered is demonstrated, not asserted: the probe "what would have happened without it?" is answered by the two previous failed attempts.
  3. The assumption challenged is the founding one. Contesting the group's own terms of reference, at the first meeting, with evidence, against immediate senior-anchored pushback — the challenge descriptor at full difficulty, not a facilitated disagreement.
  4. The compromise is owned, priced and reasoned. Timeline and totality traded; content held; the judgement stated aloud — progress was the hill; completeness wasn't. Both halves of the challenge-and-compromise descriptor, with the reasoning that resolves their tension. This paragraph is the G7 mark.
  5. The network deploys twice. The framing help for the director's revised public position — a second, quieter use, showing the network as standing capability rather than one lucky contact.
  6. The structure outlives the chair, and challenge flows back. A self-maintaining standing network resolving disputes without the candidate; a built-in norm through which the most junior members improved the output. The inclusive-environment descriptor and the G7 durability test, closed together.

Calibrating either side of G7

At SEO, architecting the multi-party structure — Version 1 plus real brokered conflict — is the scoring shape; the pre-existing network isn't yet expected. At Deputy Director, the calibration moves to networks at the most senior levels and confronting issues at board altitude. The full ladder: the Working Together pillar.

G7-specific mistakes on this behaviour

  1. The project-shaped network. Relationships created by the task and dissolved with it. G7 networks have history; show the years.
  2. Chairing narrated as collaborating. Process quality — papers, airtime, method — is the secretariat's evidence. Yours is what the relationships and the judgement changed.
  3. Challenge without discomfort. If nobody pushed back, no assumption was really challenged. Keep the resistance in.
  4. Compromise as erosion. Ground given because holding it was tiring, narrated as pragmatism. The mark needs the deliberate trade: what bought what, and why that was right.
  5. The network unaccounted for. "I knew someone who helped" without the how — the relationship's history and the mechanism of its usefulness are the evidence.
  6. Nothing survives you. If the collaboration ends when your involvement does, it was a project. The standing structure is the G7 close.

If Working Together is on your G7 advert, audit your answer for the network's history and the challenge-compromise judgement — the AI coach scores it against the real G7 descriptors and tells you which grade your telling currently sits at. Three sessions free; the G7 Complete Guide calibrates all nine behaviours to this level.

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