A fellow newsletter writer left a comment on my work recently that I haven't stopped thinking about. Proving outcomes, he said, is the part funders are asking for and the hardest part for a small charity to show. His suggestion was to build each edition around one hard number, like families housed or meals served.

He's half right. And the half he isn't right about is the most common mistake I see in funding applications and impact reports.

It's not a bad paragraph. It's clear, it's true, and it has a hard number in it. Most funders have read hundreds like it.

First, 4,212 parcels is an output. It tells the funder how busy you were. It doesn't tell them whether anything changed for anyone.

Second, the number rises when things get worse. More parcels can mean more need, more people finding you, or the same families coming back month after month because nothing else has shifted. A careful funder can read that 23% increase as a sign the work isn't working. You've handed them the doubt yourself.

Third, there's nobody in it. The only people who do anything are the volunteers, working tirelessly. The people the food bank exists for appear as a total.

That's the trouble with "one hard number" on its own. Parcels given out, meals served, hours volunteered. They're often the easiest numbers to count and the weakest evidence of change.

The question isn't "what's our biggest number?" It's "what's the number that shows something changed?"

Here's a test for any figure. Could this number go up even if nothing got better for anyone? If it could, it measures need or activity, not change. It still belongs in your application, but as context, not as your proof.

For services that respond to need, like a food bank, it's sharper still. The number can go up precisely because things are getting worse.

Even a number that sounds like a change, like families housed or people into work, needs two checks. Out of how many? And did it last?

The number that shows change is often already in your records. In this case, it's sitting in the referral list for the advice session.

But a number on its own still won't carry an application. Funders discount anecdotes, and they forget statistics. What works is the two together, plus one thing almost nobody includes.

The number shows that change happened at scale.
The person shows what that change looks like in a real life.
The limit says, honestly, what the number can't prove.

That last one feels risky. It's the part that builds trust. A funder reading an application with no limits in it assumes you either haven't checked or are hoping they won't.

Here's the same answer, opening differently. It's still a composite, and the numbers are illustrative.

Look at what's changed.

The 4,212 is still there, but it's doing a different job. It shows the scale of need, and it's no longer pretending to be the impact.

The man is the one who changes things. He asks for the side door, he comes back, he brings the letters. The volunteer mentions a session. Nobody rescues him. His story is told with his agreement and with nothing that could identify him.

The limit is one honest sentence. And the ask connects to what the number shows works. The funder isn't being asked to keep a service going. They're being asked to double the part that changes things.

The same mistake turns up in corporate impact reports, often almost word for word.

Picture a composite CSR lead at a mid-sized firm. They're writing the report that decides whether next year's volunteering budget survives. Their lead number is 1,850 employee volunteering hours, spent mentoring young people through a partner charity.

Run the test. Could that number go up even if nothing got better for anyone? Easily. Hours measure what your colleagues gave, not what changed for the young people.

The number that shows change usually sits with your partner, not in your own records. For example: of the 40 young people who had a mentor for six months, 26 had a job, an apprenticeship or a college place by the end. Ask the charity for it. Then add one young person's journey, told with their agreement, and one honest limit: the charity offers other support too, so mentoring isn't the only reason.

The reader is different. It might be your leadership team, your board or a corporate sponsor. If you're asking a company to sponsor your project, the same applies. A sponsor wants to put its name to a change, not to a count of activity. The question underneath is always the same: what did this money change?

Not every change can be counted this neatly. Some of the most important work, like prevention, belonging or feeling safe, doesn't produce a tidy number. For now, start with the work where a better number is hiding in plain sight.

If you're a charity or social impact leader, this is for you.

You can explain your work to a friend in two minutes. Then you open a funding form, or get five minutes with a donor, and it comes out flat.

Over four weeks, we'll build your core story: one clear account of what you do, what changes for people, and why it matters. It sits underneath every funding application, donor appeal and impact report, so you adapt it each time instead of starting from scratch.

Each week there's one 90-minute live session and an optional co-working hour, and the main sessions are recorded. It starts on Monday 9 November, it's live online, and there are only twelve places. It costs £295 including VAT. Three places are held at £180 for organisations with an income under £50,000 a year. If that's you, reply to this email and let me know.

Places open on Monday 19 October. Joining the waiting list is free, with no commitment, and I'll email you when they open.

This prompt looks at an impact paragraph you've already written, sorts your numbers into need, activity and change, and helps you find a better number in the records you already keep. It works for charities and for company social impact teams. It won't invent figures, it won't write the finished paragraph, and it will tell you if the number you need doesn't exist yet.

Before you paste: take out names and anything else that could identify the people you support.

Copy and paste it into Claude or Gemini.

This AI prompt is accessible to subscribers of this newsletter. It's free to subscribe, and all you need to provide is your email address ↓

Subscribe here

This is the first of four case files. I want to know if it's more useful than what I usually send. One click tells me, and there's space to add a line on why. I read every answer, and they'll decide what happens after week four.

Have a real situation you're stuck on? A grant renewal, a sponsorship proposal, a board report, a number that won't land, a story you can't find a way to tell?

Reply to this email, or message me on LinkedIn, and describe it in a few lines. I'll pick some to work through as case files in future editions.

If I use yours, I'll take out anything that could identify you, your organisation or the people you work with, and I'll show you the case before it goes out. I can't promise to use every one, but I'll read them all.

Your biggest number is rarely your best evidence. It usually tells the reader how much you did, not what changed.

Find the number that shows change. Put one person inside it. Say what it can't prove.

Then tell me what you made of this format: keep, change or drop.