How do we gather this without it taking over?

England and Wales: This page covers designing data collection in full. It is practice, not law. Scotland: the same position, with no difference in principle. Northern Ireland: the same position, with no difference in principle. The data protection side is UK-wide and is handed to other pages below.

Collecting impact data inside existing routines

The collection method that survives is the one nobody notices. It sits inside a routine that happens anyway. This step assumes you have chosen your indicators and decided whether you need a framework; it designs how each indicator gets recorded.

Take a separate process: a monthly survey, an evaluation spreadsheet, a form that lives in a different folder. That is the first thing dropped when a worker is off sick or a deadline lands. It is not dropped because people do not care. It is dropped because it is extra.

So for each indicator, find the moment that already exists:

• Counts — the attendance register, the booking system, the referral log.
• Starting points — the first session, the intake conversation, the welcome form.
• Change — the last session of a course, a review meeting, a routine check-in.
• Stories — the notes workers already write after a session.

If an indicator has no moment it fits, that is a warning. Either find one, or ask whether you need the indicator.

Designing collection backwards from the report

Design collection backwards from the report it will feed, not forwards from the programme.

We read the published reporting requirements of eight major UK funders. All eight ask a version of the first four of these, and half ask the fifth:

• what you did, and whether it was what you said you would do
• how many people, what activities, and how often
• what changed for them, in the language of change
• what the money was spent on
• what did not go to plan — four of the eight invite this in their own words

Putting the eight side by side is our own work; each funder’s requirement is its own, and binds its grant only. But the pattern is clear. If your routine record captures those things, the annual report is mostly already written.

Two practical points follow from what the funders publish. The National Lottery Community Fund and BBC Children in Need require figures, not narrative, so your routine record needs counts, not only notes. And Esmée Fairbairn Foundation accepts a report you already produce for someone else. One good record, collected once, can serve several funders.

Is there an official collection method?

No. There is no official method for how a small charity should collect its data. No regulator, funder or sector body we read prescribes one. Anyone who tells you there is a correct way is offering their own practice, however well meant.

That leaves you free to design something that fits. Keep it as small as it can be. In our view, the smallest charities need little more than a record of who they helped and what changed for them. That is this guide’s opinion, set out with its reasoning under choose your indicators.

Four questions for each indicator

The written method answers four questions for every indicator on your list:

The last question is the one people skip. Three workers keeping three spreadsheets is not a method. One shared record, in one place, is.

Test the method with the people who will run it before you finalise it. If a worker says it adds ten minutes a session, believe them and cut it down. Template: Outcomes framework (planned).

Where the data protection questions are answered

Collecting data about people raises data protection questions. Each has its own page:

• Why you may use the data at all. Using records for evaluation is a different purpose from providing the service. The lawful basis for it is settled when you get consent and handle beneficiary data ethically and establish your lawful basis for each purpose.
• What you hold and where. Add the new records to your data map.
• Who may do what with it. Your data protection policy.
• How long you keep it. Your retention schedule.
• Asking people directly. If collection includes a survey, design a beneficiary survey people will actually complete.

If you are writing a bid, the method you promise there is covered at Write the monitoring and evaluation section. Write this page’s method first and the bid section becomes a summary of it.

With the method written, the next step is the survey, if your method uses one. If it does not, go on to the step on consent and handling beneficiary data.

Common mistake: collecting what interests you, then rebuilding what the funder asks for each year

Collection is usually designed from the programme, not from the report it will feed. Design it backwards from what funders ask for, so that what you collect routinely is what you will be asked for. Our reading of eight funders’ own requirements finds the same core in all of them, and two require figures rather than narrative. Each requirement binds its own grant; the common core is our synthesis.

Common mistake: collecting once and assuming it covers delivery, fundraising and evaluation

Most small charities collect once, at intake, and reuse the data for everything. But providing a service, fundraising and evaluation are three purposes, not one. This guide’s data protection pages treat evaluation as needing its own lawful basis. That question is owned elsewhere: settle it when you get consent and handle beneficiary data ethically, before the collection method goes live.

Worked example

Saltmarsh Trust’s project workers write session notes anyway. Its funder reports are compiled from memory each year, in a rush, by the manager.

Saltmarsh designs backwards from what its funders ask: how many people, what changed for them, spend against budget, and what did not go to plan. Two of the funders it reports to want figures, so the routine record captures counts, not only notes.

It adds three things to the existing session note:

• a tick for attendance
• the two outcome questions, at the first and last session
• one line headed “what didn’t work this week”

Workers record them in the same shared file they already use.

Before any of it starts, the manager records the lawful basis for using the notes for evaluation, under the consent step.

The outcome: a one-page method, added to the existing session-note routine rather than run alongside it. It sets out what is recorded, by whom, at which session and in which shared file. Convention and funder requirement; the only legal element is the data protection one, owned by the running hub.