CS Interview Coach
Start Free

By grade

Delivering at Pace Grade 7 Example Answers (Civil Service)

Published

At Grade 7, Delivering at Pace commits the framework's most elegant inversion: the evidence that proves the behaviour at every grade below starts disproving it. Personal drive, hands-on re-planning, being the one who made it happen — SEO gold — reads at G7 as a warning light, because the descriptors have changed what pace is: ensure everyone clearly understands and owns their roles, responsibilities and business priorities, set out clear processes and standards for managing performance, deliver through providing the right resources. Pace at G7 is a property of a system you designed, running through leaders you've equipped — and the truest test of the system is what happens when you're not in the room. The panel isn't asking whether you can drive delivery. They're asking whether delivery still happens when you stop driving.

What G7 panels score on this behaviour

An operating rhythm you designed. Review cadences, milestone structures, escalation triggers with thresholds — the machinery that surfaces problems at the right altitude before they're crises. "I monitored closely" is SEO; "the weekly review escalated anything breaching X to me automatically" is G7.

Delivery through other leaders. HEOs and SEOs owning their areas with genuine authority, clear standards and the resources to meet them — your role visible in what you built and moved, not in the casework.

Resources reallocated, with the trade-off owned. Capacity moved between teams, priorities defunded to fund others — allocation decisions that made someone worse off, made by you, defended.

Performance maintained in challenging circumstances. The descriptors' phrase — meaning the system flexing under a shock without you reverting to personal heroics: the pre-built trigger firing, the contingency activating, the standard consciously adjusted rather than silently failing.

Questions this appears as

  • "Tell me about delivering a major piece of work to a demanding timescale."
  • "Describe how you've maintained delivery across teams under sustained pressure."
  • "Tell me about a time delivery was at risk and what you did."
  • "How have you built a culture of delivery in an area you led?"

Probes to expect: "What did you personally do versus your managers?", "How did problems reach you?", "What did you stop or defund?", "What happened when you were away?"

Fail vs pass at G7: the same launch

FailVersion 1 — personal drive, scores around 4/7 at G7

"I was responsible for standing up a new casework function to administer a statutory scheme with a legally fixed launch date nine months out — recruiting and training around 30 staff across three teams, building the process, and being ready for day-one demand. It was the most demanding delivery of my career and I drove it personally from start to finish. I ran the programme plan myself, chaired weekly delivery meetings across all workstreams, and kept a detailed grip on every dependency. When recruitment fell behind at month four, I intervened directly — re-prioritising the campaign with HR, sitting on sift panels myself to clear the backlog, and adjusting the training schedule to compress the lost time. When the process design hit problems in testing, I worked through the weekends with the design team to resolve the issues. I was across every detail, and it paid off: we launched on the statutory date, handled first-month demand successfully, and the permanent head of the function inherited a working operation. My director credited the launch to my personal grip and energy."

An SEO reading this would be impressed, and that's the problem. Score it against the G7 descriptors: where do other leaders own anything? Three teams exist but no team leaders appear; every intervention is personal (sift panels sat on, weekends worked, "across every detail"); problems reach the candidate by being personally gripped rather than through any designed route. The closing evidence — "credited to my personal grip and energy" — is the anti-descriptor: it proves the delivery worked and the operating model never existed. The obvious probe ("what happened when you were away?") has no good answer. A 4.

PassVersion 2 — the same launch, around 6/7

"I was responsible for standing up a new casework function — 30 staff, three teams, a statutory launch date nine months out that no one could move. My first decision was about my own role: with a legally fixed date, the single biggest risk was me becoming the throat everything had to pass through, so I spent the first month building the machine rather than driving the work. I appointed the three team leads early — before their teams existed — and gave each an owned launch area with explicit authority: recruitment and training, process design, and day-one readiness, each with a one-page remit stating what they decided and what came to me. What came to me was defined by triggers, not instinct: any milestone forecast to slip beyond ten days, any dependency between two areas the owners couldn't resolve in a week, and anything touching the statutory date — those escalated automatically through a weekly dashboard the leads maintained; everything else was theirs. The standards were equally explicit: a launch-readiness definition, agreed in month one, that every area was measured against fortnightly, so 'on track' meant the same thing in every room. The system got its test at month four, when the recruitment forecast breached the ten-day trigger — the escalation arrived through the dashboard, as designed, two weeks before it would have become visible as a crisis. My response was an allocation decision, not an intervention: I moved two experienced caseworkers from the process-design team into training-delivery roles for eight weeks — accepting a documented delay to two non-critical process refinements — and agreed with the recruitment lead a revised profile that traded five of the unfilled experienced-grade posts for eight entry-grade posts we could actually fill, with a strengthened mentoring structure to carry the experience gap. Her plan, my resourcing, both our names on the risk. What I deliberately didn't do was join sift panels or rewrite schedules — the leads did, because they'd own the consequences long after launch. I tested the system's independence directly: I took a pre-booked week's leave at month seven and changed nothing about it, and the dashboard, triggers and leads ran the fortnight's decisions without me — which I'd argue was the launch's most important test result. We launched on the statutory date, first-month demand was handled within the readiness standard, and the operating rhythm — dashboard, triggers, owned areas — transferred intact to the permanent head, who kept all three leads. Eighteen months on, the function's performance framework is still the one built in that first month, and two of the three leads have been promoted, which I take as the real verdict on giving them the launch rather than doing it for them."

The annotation:

  1. The first decision is about the candidate's own role. "The biggest risk was me becoming the throat" — the G7 thesis, stated as a design choice in month one. Version 1's energy is Version 2's named risk.
  2. Other leaders own real things, early. Leads appointed before their teams, one-page remits splitting their decisions from yours — everyone clearly understands and owns their roles delivered as architecture.
  3. Problems reach you by design. Ten-day slip triggers, unresolved-dependency triggers, statutory-date triggers, a dashboard — the escalation machinery that replaces "personal grip", and it demonstrably fires two weeks ahead of crisis.
  4. The intervention is an allocation, not a rescue. Caseworkers moved between teams, a documented delay accepted, experienced posts traded for fillable ones — resource decisions with owned trade-offs, while the leads keep the execution because they'd own the consequences.
  5. The system is stress-tested on purpose. The unchanged week of leave is the boldest sentence in the answer — the direct, verifiable claim that delivery didn't depend on the candidate. It pre-answers the killer probe.
  6. The result is the machine, not just the date. Launch met, but the emphasised outcomes are the rhythm transferring intact, the leads retained and promoted — pace as a durable property of the function, which is precisely what the G7 descriptors describe.

Calibrating either side of G7

At SEO, Version 1's directness — minus the weekend heroics, plus a real team re-plan — is close to scoring evidence; hands-on grip is still the right altitude there. Above G7, at Deputy Director, the descriptors move again: strategic priorities translated into outcome-focused objectives for managers and teams, resilience promoted across an organisation. The full grade ladder: the Delivering at Pace pillar.

G7-specific mistakes on this behaviour

  1. Heroic grip. "I was across every detail" is the anti-descriptor at G7. The panel hears a system that never existed.
  2. Invisible leaders. Teams without named leads owning named things. If no one under you decided anything, the delivery model was you.
  3. Escalation by vibe. Problems reaching you because you noticed them is SEO monitoring. Show the triggers and thresholds you set.
  4. Resource requests instead of resource decisions. Asking upward for more people is everyone's move; moving capacity between your own teams — and owning who lost out — is the G7 one.
  5. Rescue narrated with pride. If you did personally step in, frame it honestly: the trigger that should have caught it, and what you changed so the system catches the next one.
  6. No absence test. "What happened when you weren't there?" is this behaviour's defining G7 probe. The strongest answers volunteer it before it's asked.

If Delivering at Pace is on your G7 advert, test whether your answer describes a machine or a driver — the AI coach scores it against the real descriptors at G7 calibration. Three sessions free; the G7 Complete Guide runs all nine behaviours through the uncertainty principle and the delivery-through-leaders lens.

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