Corporate volunteering got bigger. Each volunteer got smaller.

Benevity, which runs a corporate giving and volunteering platform used by a large number of big employers, published its State of Corporate Volunteering 2026 using its own platform data from 2019 through 2025. The headline is genuinely good news: 23.7 million volunteer hours in 2025, a 175 percent increase since 2019. Unique volunteers rose to 1.87 million, more than triple the 2019 figure. Participation went from 10.4 percent of employees to 13.6 percent.

More people are volunteering through work than at any point Benevity has measured.

Now the number underneath. Average hours per volunteer fell from 16.4 a year to 12.7. Volunteers giving fewer than five hours a year, which Benevity calls micro-volunteers, are now roughly 60 percent of everyone on the platform.

The pool got much wider and noticeably shallower at the same time, and the second half of that sentence is the part with consequences.

Say whose numbers these are

Before going further: this is a vendor reporting on its own customers. Benevity’s figures cover employees at companies that bought a Benevity subscription and used it, which is not a random sample of US employers and is certainly not a random sample of volunteering. Companies that purchase a giving platform are companies already committed enough to purchase a giving platform.

That does not make the data useless. It is one of the only large longitudinal views of corporate volunteering anyone publishes, and the internal trend from 2019 to 2025 is measured consistently against itself. But the absolute levels should be read as “what happens on this platform,” not “what happens in America.” I would say the same about anything CECP or Points of Light publishes on their own member and applicant bases.

What a hundred people for two hours actually costs a charity

Here is the arithmetic a company almost never sees.

A hundred employees for a two-hour afternoon is 200 volunteer hours. Valued at the published US rate of $36.14 an hour for calendar year 2025, that is about $7,200 of salary-equivalent labor handed over. Which sounds like a substantial gift, and on the company’s slide it will be presented as exactly that.

Against it, the receiving organization spends: a coordinator’s time scoping work that a hundred untrained people can do safely in one afternoon; the same coordinator’s time creating that work if it does not already exist, which it usually does not; sign-in, waivers and whatever screening the roles require; a safety briefing; supervision at some ratio that keeps everyone useful; and cleanup of the parts done wrong. Then the follow-up email, the photos, the thank-you note, and a report if the company asked for impact data.

The briefing, the site walk-through and the kit hand-out cost the same whether the person stays two hours or two weeks. I learned this running camps of more than five thousand participants staffed nearly entirely by volunteers: fixed setup cost per person is the tax on every short engagement, and it does not scale down.

Now compare that to the alternative use of the coordinator’s week. In the 2023 Volunteer Management Progress Report from Tobi Johnson & Associates, gathering 1,247 responses across 36 countries, the mean volunteer-to-supervisor ratio was 21 to 30 volunteers per supervisor, the mean annual program budget in the US was $5,001 to $10,000, and 12 percent of programs ran on nothing at all. Only 20 percent of volunteer managers were fully dedicated to the work.

So the person absorbing your hundred-person afternoon is, in the typical case, part-time on volunteers, supervising several dozen of them already, and working with a budget smaller than the salary-equivalent value of the group you just sent. The gift can plausibly be negative.

Twenty percent

In the same Benevity reporting, only about 20 percent of nonprofit leaders said corporate volunteers contribute meaningfully to their long-term capacity.

One in five. From a population of nonprofits actively engaged with corporate volunteering programs, who have every commercial incentive to say something warmer.

I take that as the most important number in the report, and it is not a mystery once you have sat on the receiving end. Long-term capacity means a thing the organization can still do next quarter: a database that works, a trained cohort of tutors, a grant application that gets submitted, a rota that fills itself. Four hours of painting a fence produces a painted fence. It does not produce capacity, and the charity knows the difference even when the photographs look identical.

Why companies drifted this way

Nobody set out to build shallow programs. The drift is a measurement artifact.

Almost every corporate volunteering target is a participation rate: what share of employees took part this year. That metric improves fastest by making participation easier and shorter. Cut a role from a six-month commitment to a ninety-minute packing event and your participation rate climbs, your internal survey scores climb, and your annual report gets a better number. Nothing in the measurement system registers what the receiving organization lost.

Benevity also notes that most companies are not measuring business outcomes from volunteering at all, which means the participation rate is often the only number in the room. It is also, as the unused volunteer time off benefit shows, a number that can stay flat for years without anyone treating it as a failure. Metrics that are alone in a room get optimized, and this one has an obvious cheat.

There is a matching finding on the nonprofit side that has been sitting in the literature since 2009. Analyzing national US data from 2005 to 2007, Eisner, Grimm, Maynard and Washburn found that volunteers doing general labor came back the following year 53 percent of the time, while volunteers doing professional or management-level work came back 74 percent of the time. That data is old and I always say so, but it is the clearest evidence available that what you ask a person to do determines whether they return. I have written more about what role design does to retention. Micro-volunteering is general labor by construction. You cannot give someone a consequential role in ninety minutes.

What deeper looks like without being precious about it

I am not arguing against ever running a day of service. They are good for the company, people enjoy them, and Rodell’s 2017 work on corporate volunteering climate found that a strong volunteering climate raised commitment through collective pride among volunteers and non-volunteers alike. There is genuine internal value even in the shallow version.

The argument is that a day of service should be the front door, not the house. Four things change the picture:

Send the same people back. A group that returns quarterly to the same organization amortizes the setup cost across four visits instead of one. The second visit is worth more than the first to everyone involved, and it is the only version that produces a relationship rather than a transaction.

Send skills where the skill is the point. An accountant doing a month-end close for a charity with no finance staff, a developer fixing the donation form, a lawyer reviewing a lease. This is where the hours-to-value ratio is not just better but different in kind. It is also harder to arrange, which is why it stays rare.

Ask what they need before you decide what you are giving. Rodell’s 2021 piece in Harvard Business Review names prioritizing leadership’s preferred causes over employees’ actual interests as one of three common mistakes. The same error runs one step further out, toward the nonprofit, and I have written about what the receiving organization actually wants.

Pay for the coordination. If you are sending a hundred people, fund the staff time it takes to host a hundred people. A small unrestricted grant alongside the volunteer day is the single most welcome thing a corporate partner can do, and almost nobody does it.

The request nonprofits are making that nobody is answering

One finding in the Benevity report has almost nothing to do with volunteering and is probably the most actionable item in it. Among nonprofits surveyed, 71 percent named using AI for operational efficiency as an urgent priority. Only 3 percent use it extensively.

Sit with that gap for a second in the context of who your employees are. A company with software engineers, data analysts and operations people has, sitting in its own building, exactly the skills that gap describes. A week of one competent person’s time setting up something that drafts grant reports or cleans a donor database would land harder than a year of afternoons.

I do not know of anyone who has measured the effect of that kind of contribution on nonprofit capacity, so treat it as a reasoned bet rather than an evidenced one. But the two numbers sitting next to each other, 71 percent urgent and 3 percent doing it, are a request in plain sight.

What I cannot tell you

Whether the shift to micro-volunteering is net negative overall. It might not be. Benevity’s own participation figures show a lot of people volunteering who previously did nothing, and a first ninety minutes can lead somewhere. The Deloitte 2024 survey of 1,000 US office professionals found 90 percent went on to volunteer independently after taking part in a workplace program, though that is stated behavior in a survey rather than anything observed, and the people who say yes to such surveys are not a neutral group.

What the data does establish is that the average corporate volunteer is giving less time each year, that the receiving organizations mostly do not think this builds their capacity, and that the metric most companies report on cannot see either fact. If you run one of these programs, the honest first move is to stop reporting participation on its own and to put hours per participant next to it. That second column will be uncomfortable, which is the point of it.