Evidence a funder will accept.

The problems charities actually bring

Every funder wants a different report

A grant-funded programme usually reports to more than one body, and each one wants outcomes cut a different way. That argues for collecting once and slicing afterwards rather than running three surveys. CSV export gives one column per matrix row and a separate column for “Other” free text, which is the difference between a spreadsheet you can pivot and one you have to unpick by hand. Live charts per question, with response counts and completion, cover the quick answer at a trustees' meeting.

Partner and funder logos change per project

A charity delivering three programmes for three funders cannot have one logo bolted to the account. Collaborator logos are per survey, deliberately separate from the organisation's own brand kit, so a lottery-funded project and a council-commissioned project can carry different credits from the same workspace. Each survey also has its own accent colour.

Beneficiaries are not a captive audience

People answering a beneficiary survey are doing you a favour, often on a phone, sometimes in a waiting room. Conditional logic means they only see the questions that apply to them — someone who did not use the advice line is not asked to rate it. The conversational layout shows one question at a time and auto-advances on one-tap questions, which suits a short survey done standing up. For anything longer, save and continue issues a resume link keyed on a 24-character token.

Sensitive questions need to be asked carefully

Charity surveys often ask about circumstances people would rather not type into a random web form. Two structural answers. First, a survey only stores what you ask for: leave out name, email and address questions and there is no identifying answer in the export. Second, where you do collect personal data, it is stored in London on UK infrastructure, encrypted in transit and at rest, with row-level security throughout and organisation isolation pinned server-side. The security page has the detail, and the sub-processor list is published rather than available on request.

Volunteers are a survey audience of their own

Volunteer surveys have a different shape to beneficiary surveys: fewer respondents, more open text, and a strong need to route the answers to whoever manages that person. Roles are owner, admin and member, with multiple organisations supported, so a federated charity can keep regional workspaces separate. The last owner of an organisation cannot be removed or demoted — enforced at database level, not by a warning dialog.

What a charity survey typically looks like

A community advice charity running an end-of-programme beneficiary survey would usually build something close to this.

1. Which service did you use? A Checkboxes question using bento option tiles, each with a description line explaining what the service is, because beneficiaries do not always know the internal name for it.
2. Branching. Conditional logic routes each selected service to its own short block, so someone who used one service answers six questions rather than thirty.
3. Outcome ratings. A Matrix question with rows for confidence, knowledge and next steps, sharing one agreement scale — the format most outcome frameworks expect.
4. Change over time. A Linear scale question from 1 to 10 with labelled ends, asked about before and after.
5. Would you recommend us? A Recommend score question, 0 to 10. It is labelled “Recommend score” in the interface rather than NPS, because respondents do not speak in acronyms.
6. In your own words. A Paragraph question with a maximum length and live counter, which becomes the quote in the annual report.
7. Optional contact. An Email question, shown only if the respondent says they are happy to be contacted about a case study — a piece of conditional logic that keeps consent and collection in the same place.

Self-assessments and scored tools

A lot of non-profit work involves handing someone a structured self-assessment: a readiness check, a wellbeing baseline, a maturity model for smaller organisations you support. Scoring covers this directly. Each choice carries a score, each question carries a weight, and scoring can be switched on or off per question. Percentage-based bands are shown on the thank-you screen, so the respondent gets something back rather than a bare “thanks”. Crucially, only questions the respondent actually saw are scored — a branch that skipped ten questions does not drag the percentage down.

Running the survey

Set a close date and a response cap; both are enforced by a database trigger rather than the interface, since responses insert straight from the browser. Add a custom closed message pointing people at the next thing. Use hidden fields, declared per survey and read from the query string, to tag which outreach route a response came from — a poster QR code, a partner's newsletter, a caseworker's link — without asking the respondent to tell you. Embed mode puts the survey inside your own site rather than sending people elsewhere.

Before you publish real results, clear the test responses. That action removes uploaded files and partial responses along with the test rows, so the pilot does not quietly inflate your numbers.

Common questions

Can we credit a funder on one survey only?

Yes — collaborator logos are per survey, not per account.

Can responses be anonymous?

Yes. Nothing identifying is stored unless you ask for it or attach it through a hidden field you declared yourself.

Can respondents see their own result?

Yes, via percentage bands on the thank-you screen when scoring is switched on.

Related: local government surveys if you deliver commissioned services, healthcare surveys for patient-facing programmes, and the full list of question types. Back to all solutions.