Count What Survives: The How and Why of Measuring Social Impact

Every annual report in the social sector opens with the same number: how many people we reached. Trained. Touched. Enrolled. It's the easiest figure to collect and the most comforting to print — and it tells you almost nothing about whether a life actually changed.

A headcount is an output. It measures effort, not effect. You can train a thousand women and still leave a thousand lives exactly where you found them. The honest question isn't "how many did we reach?" It's "what was different a year after we left?"

That is the test we hold ourselves to: not the photograph at graduation, but the income that still exists twelve months later. Not attendance, but whether a woman who joined STREE is still running her business when no one is watching. Not skills certified through JIVICA, but jobs held. Not a STEAM workshop delivered in a government school, but a child who kept asking questions after we packed up. The measure of social work is what survives our absence.

Why it matters. Measuring only outputs lets you mistake activity for achievement — and it quietly misallocates everyone's money. Funders pay for reach when they should pay for change. Worse, what you measure is what you optimise: count footfall, and you will unconsciously design for footfall instead of for outcomes that last.

How we do it. It starts before the programme, not after. A clear theory of change names the outcome we're actually after, and the assumptions between our work and that outcome — so we can test each link rather than assume it. We set a baseline, because you cannot claim a change you never measured at the start. We follow up long after the certificates, because impact that can't survive a year probably wasn't impact. And we publish what didn't work, because a programme that can't be wrong can't be trusted.

None of this is harder than a headcount. It's just more honest. And honesty, it turns out, is the most innovative thing you can bring to measuring social change.

So before the next report goes out, one question is worth more than any dashboard:

If we disappeared tomorrow, what would still be standing — and are we measuring that?