Lighthouse AdvisorySLED AI Adoption Intelligence
← Back to results

From the SLED-wide archive edition of August 27, 2026

Government evaluationEffectivePublished · Jan 2026

Large workforce trial finds measurable—but smaller—time savings

UK Department for Work and Pensions · Government operations · United Kingdom

Publisher
An Evaluation of DWP’s Microsoft 365 Copilot Trial
Original publication
January 29, 2026
Source retrieved
Not recorded in the historical archive
Read original source

What happened

A 3,549-person Microsoft 365 Copilot trial paired staff feedback with econometric analysis, separating perceived gains from measured time savings.

Why it matters

The gap between reported and measured gains gives state and local agencies a realistic basis for benefits cases and broad productivity-tool rollouts.

Evidence and measured results

Ninety percent of participants reported saving time, while the econometric estimate was 19 minutes per person per day. Seventy-three percent perceived better work quality.

Limitations and uncertainty

The evaluation was not a randomized controlled trial, and several outcomes rely on participant self-reporting.

Put this evidence to work

Lighthouse Advisory interpretation, grounded in this source as summarized in the preserved archive. Enriched 2026-09-05; this does not change the original publication date. Labels below come from the analysis itself.

Sales

Role takeaway
Customer problem
agencies need a defensible benefit case before broad productivity licensing.
Stakeholders
workforce and benefits owners, collaboration-platform IT, records, privacy, and participating service managers.
Discovery
which search, drafting, or summarization tasks consume time; what baseline exists; and can saved time become useful service capacity?
Value hypothesis
a bounded copilot pilot may reduce task time with acceptable quality, informed by the trial's measured 19 minutes per day and distinct self-reported gains.
Potential engagement
readiness assessment followed by a role-based benefits pilot.
Unsupported claims
this nonrandomized UK evaluation does not guarantee equivalent U.S. agency savings, staff reductions, improved citizen outcomes, or independently measured quality improvement.

Pre-sales engineering

Role takeaway
Fit
evaluate a Microsoft 365 assistant for document-grounded knowledge work already performed in the tenant.
Architecture and integration
connect only approved information stores and preserve workflow-level telemetry alongside the current process.
Prerequisites
usable document content, permission ownership, an agreed task baseline, and output-review rules.
Constraints
local tenant configuration, licensing, records duties, and role differences limit transfer of the UK result.
Security
test retrieval against access permissions, retention, and data-loss controls before sensitive case material is available.
Proposed proof of value
compare representative task time and reviewed output quality with the existing process, record adoption separately, and distinguish observed changes from participants' perceived savings.

Delivery

Role takeaway
Work
establish role-specific baselines, correct information-access issues, configure the approved environment, train users, and review benefits before expanding licenses.
Dependencies
permission remediation, managers' participation, and access to workflow measures.
Ownership
a named benefits owner decides continuation with service managers; IT operates the platform and privacy/records teams approve handling.
Skills and adoption
provide task examples and practice in checking generated summaries and drafts.
Governance checkpoints
readiness, pilot launch, and evidence-based expansion review.
Proposed acceptance
report observed task-time change and reviewed quality against the baseline, show no unresolved access-control failures in agreed tests, and document a continue, adjust, or stop decision. Risks include optimistic self-reporting and time saved without usable service benefit.

Implementation considerations

Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.

Architecture and integration

What must connect, and where does the AI sit in the workflow?

Instrument adoption and time saved at the workflow level; connect approved copilots to governed document stores and preserve usage telemetry for benefits analysis.

Governance

Who approves, reviews and stays accountable for outcomes?

Pair licensing with role-based training, baseline measures, acceptable-use rules, and a named benefits owner who can review continuation or expansion.

Security and privacy

What data, permissions and controls need testing?

Validate tenant boundaries, data-loss controls, retention, and access permissions before assistants can retrieve or summarize sensitive case material.

The preserved archive analysis covered architecture, governance and security. Not assessed for this record: accessibility and workforce, procurement, operating model.

Publication history

  1. 2026-08-27SLED-wide archive · Issue 0110 resources
Read preserved resource versions (JSON)

Stable resource ID: dwp-copilot-trial