--- title: "Donor Retention Statistics 2026: What Each Number Actually Counts" description: "The retention rates in circulation count different things. What each one measures, the window it covers, and which can sit on the same slide." canonical: https://gratefully.io/blog/donor-retention-statistics category: "Benchmarks" date_published: 2026-11-14 date_modified: 2026-11-14 read_time: "8 min read" author: "Muddsar Jamil" keywords: "donor retention statistics 2026, donor retention rate, Fundraising Effectiveness Project, FEP retention, nonprofit retention benchmark" source: Gratefully — Donor Intelligence for Nonprofits --- # Donor Retention Statistics 2026: What Each Number Actually Counts ## Quick answer Donor retention figures in circulation count different things. FEP's 26.4% is retention year to date at the end of June 2026. Its 43.3% is a full calendar year. Virtuous reports 54.73% among its own customers. All three are accurate, and none of them is comparable with the others.
On this page - [Quick answer](#quick-answer) - [The numbers you have probably seen](#the-numbers-you-have-probably-seen) - [One sentence explains most of it](#one-sentence-explains-most-of-it) - [What each number actually counts](#what-each-number-actually-counts) - [The distinctions worth knowing by name](#the-distinctions-worth-knowing-by-name) - [One more thing that happened in 2026](#one-more-thing-that-happened-in-2026) - [What to put in your board paper](#what-to-put-in-your-board-paper) - [Checking any fundraising statistic in thirty seconds](#checking-any-fundraising-statistic-in-thirty-seconds) - [A note on where this came from](#a-note-on-where-this-came-from) - [Frequently asked questions](#frequently-asked-questions)
## The numbers you have probably seen Search for the sector's donor retention rate today and you will be given a list something like this one: - Sector average between 43.1% and 54.73% - Q2 retention slipped to 26.4% - Top-quartile organisations around 70% - First-time donor retention around 19% - New online donors retained at 24%, returning online donors at 66% - Monthly donors, 81% still giving after seven months, 71% after a year Every one of those figures is real. Every one is published by a credible organisation. They are also measuring at least four different things, and the list does not say which. If you are putting a retention number in a board paper this quarter, that matters more than the number you pick. ## One sentence explains most of it The Fundraising Effectiveness Project publishes the figure the sector quotes most. On its retention page it defines that figure in six words: > "RETENTION is computed year-to-date." A quarterly FEP report does not estimate a full year. It counts what has actually happened so far. FEP's methodology note gives the worked example: the Q2 2024 report compares January to June 2023 against January to June 2024. So when the Q2 2026 report says retention is 26.4%, it means 26.4% of last year's donors had given again by 30 June. The year was half over. The figure was never going to look like a full-year number, and it is not trying to. The last full calendar year FEP published is 2025, in its Q4 2025 report: retention "edged up from 43.1% to 43.3%". 26.4% and 43.3% are not competing estimates. One counts six months. The other counts twelve. A small detail that gives the game away: FEP labels even its Q4 figures "retained YTD". At the end of December, the year to date is the whole year. The wording never changes, only the length of the window it describes. ## What each number actually counts
FigureSourceWhat it countsWindowPanel
43.3%FEP Q4 2025All donors, all channelsFull calendar year 202515,102 organisations, 7.8M donors
26.4%FEP Q2 2026All donors, all channelsYear to date, to 30 June 202617,000 organisations, 5.7M donors
18.9%FEP Q4 2025New donors only, all channelsFull calendar year 202515,102 organisations, 7.8M donors
54.73%Virtuous 2026 Benchmark ReportDonors at organisations using VirtuousReported for 2026771 mid-sized US nonprofits
48%M+R Benchmarks 2026Online one-time donors, all2024 donors, retained during 2025Online giving only
24%M+R Benchmarks 2026New online donors2024 new donors, retained during 2025Online giving only
66%M+R Benchmarks 2026Prior online donors2024 donors, retained during 2025Online giving only
Read the panel column on its own. The 43.3% comes from a panel of 15,102 organisations. The 54.73% comes from a panel of 771. When those two appear inside one bracketed range, as they frequently do, a sector-wide measurement and a single platform's customer base have been averaged together as though they were two readings of the same instrument. ## The distinctions worth knowing by name Year to date against full year. The dominant one. Every FEP quarterly figure counts only the part of the year that has happened. Comparing a Q2 figure to a full-year figure will always show a fall, and the fall is the calendar, not your donors. This also means you should not compare FEP's Q1 to its Q2 and call the rise an improvement. There has simply been more year. The comparison FEP builds its reports around is the same quarter against the same quarter a year earlier, which is where its year-over-year numbers come from. Sector panel against platform panel. FEP draws on donation records contributed by several donor-management and online-fundraising providers, and weights the panel to correct for known imbalances against the wider sector. A vendor benchmark reports on organisations using that vendor. Virtuous is clear about this in its own write-up, which says overall retention "among the Virtuous customers in our study sits at 54.73%, nearly 12 points above the FEP average donor retention figure". That is an accurate description of what the number is, and it names the comparison openly. The label tends to come off later, once the figure is travelling on its own. All channels against online only. M+R's study covers online giving, so its figures describe a different universe from FEP's before you reach any question of window. Its definitions are worth reading slowly, because they are not symmetrical. A "new" online donor is someone who gave online in a year and in none of the previous three. A "prior" donor is someone who gave online and in at least one of the previous five. Those are two different lookbacks inside one study, chosen for good reasons that M+R explains, and neither matches how FEP counts a new donor. The windows differ too. M+R's 24% and 66% describe donors from 2024 who gave again during 2025. FEP counts all channels and runs its clock differently again. None of this makes any of them wrong. It makes them non-substitutable. M+R is explicit about one more difference. For monthly donors it does not use year-over-year retention at all. It reports the share of sustainers still giving at each month, which is why its monthly figures read as 81% at seven months and 71% at twelve. ## One more thing that happened in 2026 FEP changed its methodology this year, for the first time in five years. Its archive page states the line precisely: > "all reports prior to Q4 2025 (inclusive) use our old methodology. Starting in Q1 2026, the report is generated using a new methodology." The changes cover how the panel of organisations is built, how late-arriving donations are handled, and how the panel is weighted. FEP describes them as "refinements rather than reversals" and says the older trends have "proven directionally reliable". Two practical consequences. The first is small. FEP's own Q2 retention series runs 25.3% in 2022, a peak of 27.0% in 2024, and 26.4% in 2026. It runs straight through the methodology change without a step, so the change is not what makes the quarterly numbers look different from the annual ones. The window is. The second is worth knowing before you cite anything. 43.3% comes from Q4 2025, the last report produced under the old methodology. There is no full-calendar-year retention figure under the new one yet, and there will not be until the Q4 2026 report. Until then, the most-quoted annual number in the sector and the most recent quarterly number were produced by two different methods as well as over two different windows. ## What to put in your board paper Pick one figure, name its window, and write one sentence under it. For a full-year comparison, FEP's 43.3% for 2025 is the sector's most widely used number and the largest panel behind any of them. Write underneath it: *Fundraising Effectiveness Project, full calendar year 2025, 15,102 organisations, old methodology.* For a current reading, FEP's Q2 2026 figure of 26.4% is the most recent. Write underneath it: *Fundraising Effectiveness Project, year to date to 30 June 2026, new methodology, compared with the same point in 2025.* Do not put both on the same chart without labelling the windows. A line that drops from 43.3% to 26.4% describes a calendar, not a crisis. And if you are benchmarking your own organisation, the honest version is to compare like with like: your retention measured to 30 June against FEP's Q2 figure, or your full year against its full year. Most development teams have never run the half-year version of their own number, which is the part that makes the comparison work. Our guide to [fundraising metrics and benchmarks](/blog/fundraising-metrics-benchmarks) covers how to calculate each of these on your own file. Once you know where you stand, the question changes from measurement to action, which is what [donor retention strategies](/blog/donor-retention-strategies) is for. ## Checking any fundraising statistic in thirty seconds Three questions. Any number that cannot answer all three does not belong in a document someone will make a decision from. 1. What window? A quarter, a half year, or a full year. Year to date or annualised. 2. What panel? How many organisations, and selected how. A sector sample or one platform's customers. 3. What version? Which year's report, and whether the methodology changed between then and the figure you are comparing it to. A useful habit: whenever you find a statistic inside a report you have to fill in a form to read, ask the three questions before quoting it. Gating a report is ordinary practice and says nothing about its quality. It does mean most people repeating the number have not seen how it was built. ## A note on where this came from Everything above was read at source on 14 November 2026, on FEP's own publication site, on Virtuous's own write-up of its benchmark report, and on the M+R Benchmarks study. None of it came from a summary. That is not a boast. It is the point of the exercise. Each of these figures is accurate where it was published and becomes misleading only when it is lifted away from the sentence that described it. At Gratefully we built [cited answers](/capabilities/cited-answers) for the same reason. A number without its window is a number you cannot defend in a board meeting. ## Frequently asked questions ### What is the current donor retention rate for nonprofits? It depends on the window. The Fundraising Effectiveness Project's most recent full calendar year is 2025, at 43.3%. Its most recent quarterly reading is 26.4%, which counts donors who had given again by 30 June 2026. Both are correct, and they measure different lengths of time. ### Why do donor retention statistics vary so much between sources? Because they count different things: different measurement windows, different panels of organisations, and in some cases online giving only rather than all channels. A vendor benchmark reports on that vendor's customers, while FEP weights a sector sample. ### Is 54.73% the sector average donor retention rate? It is the figure reported in the 2026 Virtuous Nonprofit Benchmark Report, drawn from 771 mid-sized US nonprofits using Virtuous. Virtuous describes it as retention among its own customers in that study, around 12 points above the FEP figure. ### Did the Fundraising Effectiveness Project change how it calculates retention? Yes. FEP states that all reports up to and including Q4 2025 use its previous methodology, and that reports from Q1 2026 onward use a new one covering panel construction, late data and weighting. FEP describes the changes as refinements rather than reversals. --- Canonical HTML version: https://gratefully.io/blog/donor-retention-statistics More about Gratefully: https://gratefully.io/ai-instructions