.png)
UK Civil Service
Government · UK
26 min/day saved
UK Civil Service saves nearly 2 working weeks per person per year with GenAI copilots
20,000-person trial over 3 months. Average 26 minutes saved per day per civil servant.
Friction
Government organisations with high volumes of knowledge work needed productivity gains without increasing headcount.
Breakthrough
A large-scale, properly measured trial with 20,000 participants over 3 months, with a single clear metric: time spent on daily tasks.
Impact
26 minutes saved per person per day. Equivalent to nearly 2 working weeks per year. Approximately 1,130 FTE-years of capacity returned for the cohort.
Unlock the full analysis with breakthrough, impact, what made it smart and its technical approach below!
Problem
UK government organisations with large knowledge work volumes sought productivity gains and capacity headroom. The trigger: structurally more efficient operations and modernisation.
What made it smart
They ran it as a large-scale trial with a clear measurement point. This gave credible, defensible data rather than anecdotal savings claims.
Technical approach
Generative AI (including Microsoft 365 Copilot) for drafting, summarising, status updates and record updates. Framed as: a personal production line for text and knowledge work, standardisation of output formats, and governance on data and prompts.
Strategic lesson
Productivity AI delivers most value when deployed at scale with proper measurement, not as a pilot for enthusiasts, but as a standard tool for everyone.
Reflection question
If your knowledge workers got 26 minutes back every day, what would you want them to spend that time on instead?