Why a third of your volunteers don't come back next year

Every volunteer program keeps two lists without meaning to. The first is the one you recruit onto. The second is the one in your head: the people who were around last spring and aren’t anymore, and who you keep almost emailing.

The second list is usually longer than anyone wants to say out loud.

The number, and how old it is

The last time volunteer attrition was measured rigorously at national scale in the United States, the finding was that more than one third of the people who volunteer in a given year do not give any time at all the following year. The rate had gone up, from 32 percent to 36 percent between 2005 and 2006. That analysis came from the Corporation for National and Community Service and was published in the Stanford Social Innovation Review in Winter 2009, by Eisner, Grimm, Maynard and Washburn.

Of the 61.2 million Americans who volunteered in 2006, 21.7 million did not volunteer at all in 2007. That was roughly 1.9 billion hours of work that simply stopped.

I want to be straight about the age of that data, because a lot of people quote it as though it were current. It is not. It is 2006 and 2007 data published in 2009, and no comparable national study has replaced it. Treat it as the last time anyone measured this properly rather than as this year’s benchmark.

You may also have run into a claim that the average volunteer retention rate is 65 percent. I went looking for the source of that figure and could not find one. It appears on a lot of vendor blogs, each citing another vendor blog. Don’t put it in a board paper.

The finding that should change what you do on Monday

Buried in that same research is a comparison that gets almost no attention, and it is the most useful thing in the whole literature.

Volunteers doing what the study classified as general labor returned the next year 53 percent of the time. Volunteers doing professional or management-level work returned 74 percent of the time.

Same sector. Same year. Twenty-one points of difference, produced by what the person was asked to do.

This is worth sitting with, because it cuts against how most programs are built. When a coordinator gets overwhelmed, the instinct is to simplify the roles: make them smaller, make them require less handover, make them interchangeable so anyone can fill a gap on short notice. That is a rational response to a staffing problem and it is also, on this evidence, a slow way to lose people.

The volunteer who stuffed envelopes for two hours has no reason to protect the time next month. The volunteer who owns the mailing has a reason.

What “design the role up” actually looks like

I have watched this play out at fairly large scale. The international camps I have been part of run to more than five thousand participants and are staffed nearly entirely by volunteers, and the difference between the roles people come back for and the roles they quietly stop signing up for is almost never about how hard the work is. Hard work retains fine. It is about whether the role has a shape.

A role with a shape has three things a task doesn’t:

Something that is yours. Not “help in the kitchen” but “you run breakfast.” The scope can be small. It has to have a boundary that belongs to one person.

A judgment call in it. If every decision is made above the volunteer’s head, you have hired a pair of hands and they know it. Give away the small decisions: the running order, the layout, which of the two ways to do it.

A visible consequence. The volunteer should be able to point at something afterward and say what would have been different if they hadn’t been there. This is much easier for a sound engineer than for a person on a registration desk, which is exactly why the registration desk turns over faster and why it deserves more design effort, not less.

Some of the highest-retention volunteers I’ve worked with were teenagers given genuinely consequential jobs. A sixteen-year-old running a camera position for a live broadcast comes back next year. A sixteen-year-old moving chairs does not, and it isn’t because moving chairs is beneath them. It’s because nothing about it required them specifically.

Satisfaction is the leading indicator, and it moves before people leave

The best current dataset on the volunteer experience is NCVO’s Time Well Spent 2023, based on YouGov fieldwork with 7,006 UK adults in late 2022. It found a steep gradient between how satisfied volunteers were and whether they expected to keep going:

How satisfied they were	Likely to continue
Very satisfied	90%
Satisfied	81%
Dissatisfied	26%

Look at the middle row. “Satisfied” — the answer most of us would be delighted to see on a survey — is already nine points down from “very satisfied.” The gap between fine and good is doing real work here.

The same survey found overall satisfaction had slipped from 96 percent in 2018 to 92 percent in 2023, that 26 percent of volunteers said volunteering felt too much like paid work (up from 19 percent), and that 24 percent reported unreasonable expectations of them (up from 17 percent). Those three moves are the same story told three ways.

Nobody hands in a resignation letter

Here is the operational problem. In that NCVO data, 20 percent of people who had stopped volunteering gave no reason in particular. Not a complaint. Not a falling-out. The largest single named reason was having less time because their circumstances changed, at 37 percent. Only 10 percent said they were unhappy with the organization or its management.

Which means your exit interviews, if you do them, are going to be uninformative most of the time. People don’t quit volunteering. They get busy, they miss one, missing the second one is easier, and four months later they are no longer a volunteer without ever having decided to stop.

That has a practical consequence: the first missed shift is the event worth reacting to, not the third. By the third you are not preventing a departure, you are attempting a recovery, and those are very different conversations.

Four things worth doing this month

Count it before you fix it. Take last year’s volunteer list and this year’s, and calculate what share came back. Most programs have never done this and are startled by the answer. You cannot tell whether anything you change is working without this number.

Sort your roles by how replaceable the person is. Anywhere you can honestly say “anyone could do this,” you have found a retention risk. Not a role to delete — a role to redesign.

Set a rule for the first miss. Someone contacts them within 48 hours, and it is a human question, not a rota question. “Everything alright?” beats “can you do the 14th?”

Ask the satisfied people what would make it better. The dissatisfied ones have mostly gone already. Your gains are hiding in the gap between satisfied and very satisfied, and the people in that gap are still in the room to be asked.

What this evidence can and can’t tell you

The role-design finding is a difference between groups, not a controlled experiment. Nobody randomly assigned volunteers to envelope-stuffing and to management work. It is plausible that the kind of person who takes on a professional-level role was more committed before they arrived, and the number would overstate what redesigning a role does for you.

I still think it’s the most actionable finding available, for a simple reason: the alternative explanations mostly still point at the same action. If committed people gravitate to consequential roles, then having consequential roles available is how you keep committed people. Either way you end up designing the work.