03 · Annual Impact Report Template

Format: 24-page A4 PDF + web HTML version + plain-text summary. Owner: Head of Impact. Cadence: Published within 90 days of MEM's financial year-end. Audience: Public.

The annual report is the public version of what we already tell every client and the board. If we wouldn't put a number in front of a journalist, it doesn't go in.

Structure

Section A — What we said we'd do (pp. 1–3)

• Recap of last year's report commitments.
• Score each: done / partially done / not done / dropped (with reason).
• Owner-named accountability table.

Section B — Who we worked with (pp. 4–6)

• Number of client orgs, by sector and size band. Not naming clients without consent.
• Clients who consented to be named (logo wall + one-line attribution each).
• Cohorts delivered, hours, participants reached.
• Geographic spread (UK nation + region; major cities + rural split).

Section C — What changed (pp. 7–14)

Rolled-up across all clients, using the four-tier framework (file 01).

• Outputs: totals.
• Outcomes: weighted-average deltas, with n and confidence per metric.
• Impact: aggregate where clients consented to share HR data; otherwise narrative.
• SROI: rolled-up ratio with confidence factor and methodology link.

Section D — Where we got it wrong (pp. 15–17)

This section is mandatory and minimum 2 pages. If it's shorter we haven't looked hard enough.

• Cohorts that underperformed and why.
• Corrections issued (file 10) — date, what was wrong, what we changed.
• Coach exits during the year (count only, not names; file 09 of Sprint 7).
• Safeguarding patterns (anonymised, aggregated only) that shaped how we changed the programme.

Section E — Our coach bench (pp. 18–20)

• Active coach count, by accreditation level (Sprint 7 / 06).
• Recruitment funnel actuals vs targets (Sprint 7 / 02).
• Diversity actuals vs targets (Sprint 7 / 02). Where we missed, what we're doing.
• CPD hours delivered and supervision attendance.
• Average QA score and re-observation triggers.

Section F — Methodology & assurance (pp. 21–24)

• The framework (file 01) summarised.
• SROI methodology (file 04) summarised with link to full method.
• Confidence factor explanation (file 05).
• Assurance statement from external assurer (file 07) — verbatim, no edits.
• Suppression statistics: how many metrics were suppressed and why.
• Commitments for next year, with owners and deadlines.

Reproducibility commitment

Published alongside the report PDF on the public site:

• Anonymised aggregate dataset (CSV) backing every number.
• SROI calculation spreadsheet with every assumption surfaced.
• A README explaining how to recompute the headline figures.

What we will not include

• Any number we can't reproduce from the dataset.
• Named participants, ever.
• Named coaches without their written consent for the year.
• Named clients without their written consent for the year (and the wording approved by them).
• Industry rankings or comparisons unless source-cited and current.
• Hero-shot stock photography. Anonymised, illustration-led design instead.

Versioning and corrections

• Versioned (v1.0 at first publish; v1.1 etc. for any correction).
• Changelog page kept in the web version.
• Withdrawn numbers struck through, not deleted — with a note pointing to the corrected version.
• Correction policy: file 10.