The definitions, without the jargon
An output is what you did or produced. Sessions run. Parcels handed out. Calls made. People through the door. It is a record of activity, and it is countable at the moment it happens.
An outcome is what changed for someone because of that activity. Somebody knows something they did not know, can do something they could not do, feels differently, or is in a different situation. Outcomes are about people, not about your diary.
Both matter. A report with no outputs is vague; nobody can tell whether you delivered anything. A report with only outputs is busy; nobody can tell whether it mattered. The two do different jobs and you need both.
Why this trips everybody up
Outputs are generated by the work itself. If you run a session, the session is recorded. No extra effort, no awkward questions, no chasing people. They also feel like progress, because they are progress, a busy month genuinely looks better than a quiet one.
Outcomes take a deliberate act. Somebody has to ask a person how things are, and ask again later, and write both down in a way that can be compared. That is a second job on top of delivery, usually falling to whoever is already doing three jobs. So it slips, and the report ends up counting sessions.
Nobody sets out to report only activity. It happens because activity records itself and change does not.
Meanwhile the reader is asking a different question. A funder deciding whether to renew, a commissioner reviewing a contract, a trustee signing off an annual report, a member deciding whether to keep giving, all of them want to know whether anything is different for the people you work with. The output number tells them you were busy. It does not answer their question.
The causal chain, and its honest caveat
An output is supposed to lead to an outcome. In a decent programme it usually does. But the arrow between them is a claim, not a fact, and the arrow is where reports get into trouble.
Output
12 employment workshops delivered, 48 attendances recorded.
Leads to, ideally
Someone attends, gains skills or confidence, and behaves differently afterwards.
Outcome
Participants apply for roles they would previously have ruled out.
Caveat worth stating in your reports
Showing that an output happened does not prove it caused the outcome. Two things moving together is not the same as one causing the other. Someone’s circumstances may have improved because of a new job, a change at home, or another service entirely. Software cannot settle that question and neither can a spreadsheet establishing cause needs evaluation design, and often a comparison group.
What you can defensibly show is what you delivered, to how many people, what evidence you collected, from whom, and what changed over that period. Stated carefully, that is strong. Stated as proof of cause, it is an overclaim waiting to be picked apart.
Six worked examples
The same distinction, across different kinds of delivery. Notice that in every row the left-hand column can be counted from your own records, and the right-hand column requires you to have asked somebody something.
| Output: what you did | Outcome: what changed for someone |
|---|---|
| Youth work36 after-school sessions run, average attendance 14. | Young people say they have somewhere to go where they are known by name. |
| Food bank1,400 emergency parcels issued to 310 households. | Households report fewer days in the month with no food in the house. |
| Elderly support780 home visits and 1,100 welfare calls completed. | Older people describe speaking to someone most weeks rather than most months. |
| Employment support12 workshops and 90 one-to-one advice appointments. | Participants apply for roles they previously ruled themselves out of. |
| Mental health220 peer support group places filled across the year. | People say they can recognise a dip early and know what to do about it. |
| Education48 one-to-one reading sessions delivered in two primary schools. | Children choose to read outside lessons and say they find it easier. |
How do you actually evidence an outcome?
The underlying principle is simple, and it is the same one whether you use paper, a spreadsheet or software. To show that something changed, you need a picture of how things were before, and a picture of how things were later, recorded in a way that lets you compare the two.
In practice that means asking the person a small number of questions near the start, how they feel about something, what they can currently do, what their situation is, writing the answers down, and then asking the same questions again after a sensible period. The comparison is the evidence. Without the first conversation you have an impression at the end and nothing to hold it against.
Making that workable
- Ask few questions. Two or three you will actually repeat beat a fifteen-question form completed once and abandoned.
- Ask the same questions both times, in the same words. Changing the wording quietly destroys the comparison.
- Pick a sensible gap. Long enough for change to be plausible, short enough that people are still around.
- Record who did not answer. Coverage is part of the finding, not an embarrassment to hide.
- Keep the person's own words where you can. A sentence someone said carries weight a scale never will.
Two honest limits. First, self-reported change is genuinely useful but it is not neutral: people tend to be generous about a service run by someone sitting in front of them. Say so in the report; it reads as competence, not weakness. Second, a before-and-after comparison shows that something changed over the period. It does not on its own establish that you caused it. Describe your contribution and leave the causal claim alone.
Where a record exists elsewhere, school attendance, tenancy sustained, an appointment kept, that is worth having alongside what people tell you. Two independent sources pointing the same way is a much stronger position than either one.
If you have far more outputs than outcomes
You almost certainly do. Most small organisations have years of attendance figures and a handful of quotes. That is normal, and it is not a failure. It is what happens when a small team spends its time delivering rather than measuring, which is the right priority most weeks.
It does tell you where the next bit of effort belongs. Pick one outcome, the one your board or your next application most needs, and start asking two questions about it, at the start and at the end. In six months you will have something you have never had before, and you will not have added an hour a week to anybody’s workload.
If you have not written down why you expect your activity to produce that outcome, the theory of change guide is the place to start, and what funders want in an impact report covers how to present what you end up with.
ImpactProof records the evidence as you deliver and keeps every figure linked to what it came from, so the outcome column stops being the one you fill in last.
