Sponsor Impact One-Pager
When To Use
The sponsor has five minutes and you have forty pages of evidence. The programme has run, the figures are in, and the person who signed the agreement needs to know whether it worked and what you want from them. It answers one question: against what we agreed, what happened, and what do you need to decide?
When Not To Use
If the programme is still running and an indicator is slipping, send a Programme Early Warning Note now; do not wait for this page. For the executive team and finance, who did not sign the agreement and want the cost and the hard questions, use Executive Impact Briefing; for the whole function across a quarter, use Quarterly L&D Report.
Inputs
- Your Success Measures Agreement, with the measures and what "worked", "partly" and "did not" mean
- Your Before-and-After Results Table, and the contribution analysis if you ran one
- The barrier list from the cases and the comment themes, if you have them
If you have none of this, I start from the measures and figures you can paste, mark the page "no agreement on file: first draft", and leave every cell I cannot fill blank. Claude Docs (beta) gives you a page to edit and share; if you also need the Executive Impact Briefing, write both in the same conversation so the figures match.
Approach
Bottom line up front, from the US Army's writing standard (Army Regulation 25-50): the main point at the beginning, active voice, understood in a single rapid reading. It is paired with the CDC guide "Evaluation Reporting: A Guide to Help Ensure Use of Evaluation Findings": define the audience, the action you want from it and what lies within its influence, in plain language. The judgement is in what you leave off the page, and in where the awkward caveat goes. The failure it prevents: a page that opens with the method and the response rate, reaches "partly" at the bottom, and puts the one caveat that changes the conclusion in a footnote the sponsor never reads.
Workflow
- Ask three questions at once: which decision do you need from the sponsor (continue, change, stop or scale), and by when? What lies within this sponsor's power to change? What does the sponsor believe today about how it went?
- Set each measure against the agreement's verdicts, in the sponsor's own terms, and state the overall verdict: worked, partly or did not. If the agreement does not say how to combine measures, the page reports them one by one and says the overall verdict is yours to state.
- Write the first two lines: the verdict against the agreement, with the caveat that would change the reader's conclusion, then the decision asked and its date. No softening sentence comes first.
- Build one table in the agreement's order: agreed measure, what was agreed, what happened, claim strength (shown, suggested, unknown), source and voucher. Figures are copied from the results table, never retyped from memory or rounded differently. A group suppressed there stays suppressed here.
- Write three short lists: what changed, what did not, what is unknown. "Unknown" is never left off, and each unknown says what it would take to know.
- Add what helped and what blocked, from the cases and the comment themes, each with the role that could act on it. Put first what the sponsor can change. No person or team is singled out as the reason for a result.
- Cut to one page. Everything else is named in an annex list. You read it against the sources, and you send it.
Output Format
# Sponsor Impact One-Pager
Programme: [name] | For: [sponsor name] | From: [name, role] | Date: [date] | Agreement signed: [date, or "no agreement on file: first draft"]
**Answer:** [worked / partly / did not] against the agreement. [The caveat that would change this conclusion.]
**Decision asked:** [continue / change / stop / scale], by [date].
## Measure by measure
| Agreed measure | What was agreed | What happened | Claim strength | Source and voucher |
|---|---|---|---|---|
| [measure, in the agreement's order] | [the sponsor's terms] | [your figure, group size] | [shown / suggested / unknown] | [source, name] |
## What changed, what did not, what is unknown
- Changed: [measure and figure, from the table]
- Did not change: [measure and figure, from the table]
- Unknown: [what, why, and what it would take to know]
## What helped and what blocked
| Helped or blocked | What (a condition, never a person) | From | Role that could act |
|---|---|---|---|
| [helped / blocked] | [condition] | [cases / comment themes] | [role] |
## Annex
- [Document, where it is, who holds it]
## Decision
[Sponsor name] decides by [date] whether to continue, change, stop or scale [programme], and names who owns each barrier to be removed.
Done When
- The first two lines carry the verdict, its main caveat and the decision asked with a date
- Every agreed measure has a row, in the agreement's order, including the ones that did not move
- Each figure matches the results table to the digit, with its source and voucher
- The "unknown" list is filled in, and the page fits on one page
Quality Bar
- Write "the figures show" for shown, "the evidence suggests" for suggested, "we do not know whether" for unknown
- A measure with no data gets a row that says so; it is not dropped to tidy the page
- The same figure never differs between this page, the briefing and the quarterly report
- No benchmark, comparison or quote is added to make the page land
- The page says what the evidence shows and what it does not; no figure appears without its source, and the sponsor decides
Next
Run impact-early-warning-note (Programme Early Warning Note) to agree the triggers for telling this sponsor early next time.
About the makers
This pack is made by 90n (90n.ai): AI-native ways of working and tools for Product, Marketing and HR teams, built by ex-McKinsey founders. We map where time and money leak, redesign the work around AI and stay until your people run it themselves, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).
1---2name: impact-sponsor-impact-one-pager3description: Drafts a Sponsor Impact One-Pager from evidence you already hold, producing the answer and the decision asked in the first two lines, a table of each agreed measure against what happened, and short lists of what changed, what did not, what is unknown, what helped and what blocked. Use for "run impact-sponsor-impact-one-pager", "one page summary for the sponsor", "training impact report", "training evaluation report", "report training results to the sponsor", "executive summary of training results", "bottom line up front", "what do I tell the sponsor", part of the Claude Guide for Training Impact Pack by 90n.4metadata:5 source: https://skillmd.com/skills/polar-bear-org/impact-sponsor-impact-one-pager6---78# Sponsor Impact One-Pager910## When To Use11The sponsor has five minutes and you have forty pages of evidence. The programme has run, the figures are in, and the person who signed the agreement needs to know whether it worked and what you want from them. It answers one question: against what we agreed, what happened, and what do you need to decide?1213## When Not To Use14If the programme is still running and an indicator is slipping, send a Programme Early Warning Note now; do not wait for this page. For the executive team and finance, who did not sign the agreement and want the cost and the hard questions, use Executive Impact Briefing; for the whole function across a quarter, use Quarterly L&D Report.1516## Inputs17- Your Success Measures Agreement, with the measures and what "worked", "partly" and "did not" mean18- Your Before-and-After Results Table, and the contribution analysis if you ran one19- The barrier list from the cases and the comment themes, if you have them20If you have none of this, I start from the measures and figures you can paste, mark the page "no agreement on file: first draft", and leave every cell I cannot fill blank. Claude Docs (beta) gives you a page to edit and share; if you also need the Executive Impact Briefing, write both in the same conversation so the figures match.2122## Approach23Bottom line up front, from the US Army's writing standard (Army Regulation 25-50): the main point at the beginning, active voice, understood in a single rapid reading. It is paired with the CDC guide "Evaluation Reporting: A Guide to Help Ensure Use of Evaluation Findings": define the audience, the action you want from it and what lies within its influence, in plain language. The judgement is in what you leave off the page, and in where the awkward caveat goes. The failure it prevents: a page that opens with the method and the response rate, reaches "partly" at the bottom, and puts the one caveat that changes the conclusion in a footnote the sponsor never reads.2425## Workflow261. Ask three questions at once: which decision do you need from the sponsor (continue, change, stop or scale), and by when? What lies within this sponsor's power to change? What does the sponsor believe today about how it went?272. Set each measure against the agreement's verdicts, in the sponsor's own terms, and state the overall verdict: worked, partly or did not. If the agreement does not say how to combine measures, the page reports them one by one and says the overall verdict is yours to state.283. Write the first two lines: the verdict against the agreement, with the caveat that would change the reader's conclusion, then the decision asked and its date. No softening sentence comes first.294. Build one table in the agreement's order: agreed measure, what was agreed, what happened, claim strength (shown, suggested, unknown), source and voucher. Figures are copied from the results table, never retyped from memory or rounded differently. A group suppressed there stays suppressed here.305. Write three short lists: what changed, what did not, what is unknown. "Unknown" is never left off, and each unknown says what it would take to know.316. Add what helped and what blocked, from the cases and the comment themes, each with the role that could act on it. Put first what the sponsor can change. No person or team is singled out as the reason for a result.327. Cut to one page. Everything else is named in an annex list. You read it against the sources, and you send it.3334## Output Format35```markdown36# Sponsor Impact One-Pager37Programme: [name] | For: [sponsor name] | From: [name, role] | Date: [date] | Agreement signed: [date, or "no agreement on file: first draft"]38**Answer:** [worked / partly / did not] against the agreement. [The caveat that would change this conclusion.]39**Decision asked:** [continue / change / stop / scale], by [date].40## Measure by measure41| Agreed measure | What was agreed | What happened | Claim strength | Source and voucher |42|---|---|---|---|---|43| [measure, in the agreement's order] | [the sponsor's terms] | [your figure, group size] | [shown / suggested / unknown] | [source, name] |44## What changed, what did not, what is unknown45- Changed: [measure and figure, from the table]46- Did not change: [measure and figure, from the table]47- Unknown: [what, why, and what it would take to know]48## What helped and what blocked49| Helped or blocked | What (a condition, never a person) | From | Role that could act |50|---|---|---|---|51| [helped / blocked] | [condition] | [cases / comment themes] | [role] |52## Annex53- [Document, where it is, who holds it]54## Decision55[Sponsor name] decides by [date] whether to continue, change, stop or scale [programme], and names who owns each barrier to be removed.56```5758## Done When59- The first two lines carry the verdict, its main caveat and the decision asked with a date60- Every agreed measure has a row, in the agreement's order, including the ones that did not move61- Each figure matches the results table to the digit, with its source and voucher62- The "unknown" list is filled in, and the page fits on one page6364## Quality Bar65- Write "the figures show" for shown, "the evidence suggests" for suggested, "we do not know whether" for unknown66- A measure with no data gets a row that says so; it is not dropped to tidy the page67- The same figure never differs between this page, the briefing and the quarterly report68- No benchmark, comparison or quote is added to make the page land69- The page says what the evidence shows and what it does not; no figure appears without its source, and the sponsor decides7071## Next72Run impact-early-warning-note (Programme Early Warning Note) to agree the triggers for telling this sponsor early next time.7374## About the makers7576This pack is made by 90n ([90n.ai](https://90n.ai)): AI-native ways of working and tools for Product, Marketing and HR teams, built by ex-McKinsey founders. We map where time and money leak, redesign the work around AI and stay until your people run it themselves, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).