A volunteer survey that tells you something

The most common volunteer measurement activity in the sector is the annual feedback survey. Seven in ten programs run one, according to the 2023 Volunteer Management Progress Report from Tobi Johnson & Associates. It is by some distance the most reported thing on their list.

It is also, in most of the programs I have seen, the least consequential. The survey goes out, the answers come back overwhelmingly positive, the results get summarized in a paragraph for the board, and nothing about next season is different because of it.

That is not an argument against surveying volunteers. It is an argument against the survey most people are sending.

Ask about four things, not fourteen

The starting point I would use is the Volunteer Satisfaction Index, published by Galindo-Kuhn and Guzley in 2001. Its value is not that it is fashionable. It is that somebody already did the work of establishing which dimensions of volunteer satisfaction hang together, so you do not have to invent question categories over coffee.

Four dimensions:

Organizational support. Does the volunteer feel the organization gives them what they need to do the job? Training, information, someone to ask, equipment that works.

Participation efficacy. Do they believe their effort makes a difference to something outside themselves? This is the one that dies quietly when a role has no visible consequence.

Empowerment. Do they have any say in how their own work is done? Freedom to make the small decisions inside their patch.

Group integration. Do they feel part of the group? Whether they know people’s names and whether people know theirs.

Two or three plain questions each gets you a survey of ten or twelve items, which someone will finish on a phone in a parking lot. Every question you add past that trades a small amount of information for a larger amount of abandonment.

Write the questions in the language your volunteers use, not the language of the index. “I have what I need to do my role well” beats anything containing the word “organizational.”

The question that carries the most weight

Add one more: how likely are you to recommend volunteering here to a friend, on a scale of zero to ten.

That is Net Promoter Score, borrowed from customer research, and I was skeptical of it in this setting for years. Piatak and Carman used it as a volunteer outcome measure in their 2023 study of 323 volunteers and it behaved sensibly, moving with the things you would expect to matter. Their strongest predictor of a volunteer recommending the organization was appropriate training, ahead of the logistics that most programs work hardest on.

The reason I like the recommend question specifically is that it is behavioral in a way satisfaction is not. Saying you are satisfied costs nothing. Saying you would put your friend’s Saturday into this organization’s hands is a claim with something at stake, and people answer it more honestly.

Do not average the answers

Here is where most survey reporting goes wrong, and it is worth being blunt about it.

NCVO’s Time Well Spent 2023, with YouGov fieldwork among 7,006 UK adults, asked volunteers how satisfied they were and how likely they were to keep going. Ninety percent of the very satisfied expected to continue. Eighty-one percent of the merely satisfied. Twenty-six percent of the dissatisfied.

Look at the middle. “Satisfied” is the answer that makes a board paper read well, and it is already nine points down from the top box. The distance between fine and good is where your attrition is being manufactured.

So report the share who chose the top box, and track that share year over year. A mean score sitting comfortably above “good” tells you nothing about whether you have a room full of enthusiasts or a room full of people who are perfectly content to be somewhere else next spring. The same NCVO study found overall satisfaction had slipped from 96 percent in 2018 to 92 percent in 2023, a move that would be invisible inside almost any averaged score.

The people you are not asking

This is the trap, and nearly every program falls into it.

A survey sent to your current volunteer list is a survey of the people who stayed. The ones who would have told you something useful stopped coming in February and are not on the list anymore. Your results are a photograph of the survivors, and survivors are cheerful.

NCVO found that 20 percent of people who had stopped volunteering gave no reason in particular for stopping. Not a complaint, not a falling-out, nothing you could have escalated. Most leaving is drift rather than grievance. Which means it will never appear in a survey of the current rota, because by the time someone drifts they have stopped opening your emails.

Two fixes, and neither is expensive. Send the annual survey to everyone who volunteered in the last eighteen months, not to the active list, and report the two groups separately. The gap between them is the most informative number the whole exercise will produce. Then, separately, survey new volunteers after their first shift while they still answer, because that group has the highest attrition and the shortest memory of what you actually did to them.

I have run volunteer programs where the people involved were spread across around eighty countries, and the thing that surprised me most was not translation. It was that the response rate from people who had done one shift was a fraction of the rate from people who had done twenty, and we kept designing improvements from the answers of the twenty-shift group. We were listening carefully to the people who had already decided to stay.

The two questions to send after a first shift

The first-shift check is short enough that people answer it on the bus home, and it should ask only what you can act on before their second shift.

The first question: did you have what you needed to do the job today. That is the organizational support dimension in ordinary words, and it maps onto the strongest finding in the Piatak and Carman data, which was that appropriate training predicted both recommending the organization and intending to return. A first-timer who says no is telling you your onboarding failed, and you have about a week to fix it before they conclude they are not the sort of person who does this.

The second: was there anything you were unsure whether you were allowed to do. That one sounds odd until you start reading the answers. It surfaces the boundary problems that people never raise as complaints, because nobody complains about being unsure. They just do less next time.

Then call the ones who answered badly. Not email. A phone call from a person who can change something, inside the same week.

Timing, length, and what to do with the answers

Once a year for the full survey. Two questions after a first shift. That is the whole cadence, and I would resist anything more.

The argument for surveying quarterly is that you catch problems earlier. The argument against is that the same NCVO research found 26 percent of volunteers said volunteering felt too much like paid work, up from 19 percent in the previous wave, and 24 percent reported unreasonable expectations of them, up from 17 percent. A quarterly survey obligation is a small unpaid administrative task you are handing to someone who came to help at a food bank. Do it once, do it well.

Length: under five minutes, honestly measured by timing yourself. Include one open text box asking what would make this better, and read every answer personally rather than running them through a summarizer. In the Piatak and Carman data, 13.4 percent of open-ended improvement suggestions asked the organization to be more inclusive, which is the kind of finding that only survives if a human reads the raw text.

Then the part that decides whether any of this was worth it. Within two weeks, write one page back to the volunteers: here is what you told us, here are the two things we are changing, here is one thing we are not changing and why. The last item is the one that builds credibility. A program that only ever reports the changes it made looks like a program that asked leading questions.

If you cannot commit to that two-week reply, do not send the survey. An unanswered survey is worse than no survey, because it teaches people that telling you things does not work.

One thing to check before you write a single question

Ask whether you reimburse expenses and whether your volunteers know it. NCVO found only 55 percent of organizations reimburse expenses, 16 percent of volunteers were unsure what the policy even was, and 20 percent of young people named expenses as a barrier to volunteering at all.

That is a question you can answer today without surveying anyone, and it will do more for your under-25 retention than any amount of satisfaction data. If you want the wider case for keeping the measurement list short, I made it in which volunteer numbers are worth collecting.