Donor Segmentation That Actually Moves Money

"Segmentation" is one of those words charities have been using for so long that it has stopped meaning anything. Most fundraising teams have a segmentation. Most of those segmentations sit in a slide deck. The CRM, meanwhile, sends the same email to everyone.

The question is not whether you have a segmentation. It is whether your segments do anything. By "do anything" I mean: does each segment receive a different message, in a different cadence, with a different ask? If the answer is "no, but the report colours are different," you have a taxonomy, not a segmentation.

The trap of the over-built segmentation

In my experience, the most common failure mode is too many segments. A team I worked with last year had 47 active audience codes in their CRM. When we mapped which ones drove different content, the answer was four. The other 43 segments were just labels - historic campaign codes, legacy persona names, leftovers from a long-dead consultant project.

Forty-three segments that nobody uses are not "fine because we have them." They are tax. Every time a fundraiser plans a campaign, they have to navigate around them. Every time data quality drifts, they hide it. Cut them.

Start with RFM, finish with RFM+

The simplest, most durable model is Recency, Frequency, Monetary value. Three numbers per donor. From those, you get six tiers that cover most of what fundraising teams actually want to do.

The six core segments

1. Active high-value: gave in the last 12 months, multiple gifts, top quintile by total.
1. Active mid-value: gave in the last 12 months, regular cadence, middle quintile.
1. Active low-value: gave once or twice in the last 12 months, low gift amount.
1. Lapsing: 12–24 months since last gift.
1. Lapsed: 24–48 months since last gift.
1. Cold: 48+ months since last gift; possibly archive candidates.

That is six segments, queryable in any CRM in a single SQL statement or filter view. Each one earns its own ask cadence and content angle. That is a working segmentation.

The "+" - the strategic overlays

Once RFM is in place, layer two or three lightweight overlays:

• Channel preference (email-engaged vs. direct-mail-engaged vs. multi).
• Cause affinity, where you genuinely know it (programme attendance, restricted gift history).
• Major-gift potential, if and only if you have a real prospect researcher reviewing it.

Three overlays times six core segments would in theory give you 18 buckets, but in practice you will only ever message 8–10 of them differently. That is fine. The model is the map, not the territory.

Translating segments into action

A segmentation that does not change a campaign plan is decorative. The way to test yours is to take your last big appeal and ask: "What did this segment receive that the segment above and below them did not?"

Real differences look like this:

• Active high-value: hand-signed letter, named fundraiser sign-off, no online ask in the body.
• Active mid-value: full-colour pack, mid-tier ask amount, online and post in parallel.
• Lapsing: a "we miss you" piece with a smaller, lower-friction ask.
• Lapsed: a re-permission email and an offer (free guide, supporter survey) before any ask.
• Cold: archived, with a once-a-year reactivation email - measured ruthlessly.

If those differences are not visible in your campaign plan, your segmentation is not finished - your campaign is.

Common mistakes - and how to dodge them

Mistake 1: Segmenting only by what is convenient

A lot of charity segmentations are built from whatever fields the CRM had spare. Postcode area, age band, original acquisition source. These can be useful overlays, but they are weak primary segments. Behaviour beats demographics. Always.

Mistake 2: Hard-coding segments into campaign tools

I have seen segments built directly into a marketing automation tool, where the only person who knows the rule has left. Segments belong in the CRM (or the warehouse). Marketing tools should consume them, not own them.

Mistake 3: Not refreshing the lapsed line

A "lapsed" segment that hasn't been recalculated for six months is including people who gave last month. You will end up sending a "we miss you" email to your most active donor. It happens. It costs trust. Refresh nightly if you can.

What good looks like, in one sentence

Six core segments, three overlays, refreshed nightly, owned by one person, written into the campaign plan, reviewed quarterly.

That is it. Anything more elaborate has to earn its place by changing a fundraiser's decisions. Anything less, and your CRM is a list, not a strategy.

A 30-day rollout plan

1. Week 1: Audit current segments. List them. Map which ones drive different content. Mark the rest for archive.
1. Week 2: Build the six RFM segments. Save the queries. Document the rules in plain English.
1. Week 3: Pick one upcoming campaign. Plan it segment-by-segment, even if the differences are small.
1. Week 4: Run the campaign. Compare segment-level response. Decide what overlay to build next.

Thirty days. Six segments. One signed-off rule book. That is a segmentation that will hold up under pressure - and quietly move money in the right direction.

Further reading

Lead-Scoring for Charities, Without the Hype | Choosing a Charity CRM in 2026 | The Five-Minute CRM Health Check