How to evaluate recurring-giving conversion claims from fundraising vendors

Fundraising vendors compete on one headline number: how many one-time donors they turn into recurring givers. The number is easy to print and hard to verify. Before you sign, you need to know whether a claimed result reflects donations the vendor actually caused, or activity that would have happened anyway.

This guide shows how to read those claims, where the common ones mislead and what to ask before money changes hands.

What counts as a real recurring-giving conversion claim?

A real claim measures incremental lift: the extra recurring donors or revenue that exist because of the vendor, compared with a group that did not get the treatment. It is proven with a control group, not inferred from totals.

Everything else is a proxy. Proxies can be useful context, but they do not tell you what the vendor added.

The three types of claims, and what each one proves

Most vendor numbers fall into one of three buckets. They are not equally trustworthy.

Incremental lift

This is the standard to hold out for. The vendor runs the treatment on part of your file and holds back a comparable control group. The difference between the two groups is the lift. Because both groups face the same season, economy and appeal calendar, the gap is a fair estimate of what the vendor added.

Raw activity

Raw activity counts conversions during a period and attributes them to the tool. The problem is attribution. A share of one-time donors convert to recurring giving on their own, through your other appeals or through loyalty built over years. Raw activity claims quietly fold those donors into the vendor's result.

Platform-wide benchmarks

Benchmarks describe an average across many organizations. Your donor file, cause, country and data quality are not the average. A benchmark can set rough expectations, but it is not evidence of what you will get.

Why raw activity and benchmarks inflate results

The gap between a raw number and a tested number is the baseline conversion rate: the recurring givers you would have won with no vendor at all.

Say 1,000 donors see a vendor's upgrade flow and 60 convert. The vendor reports a 6% conversion rate. If a held-out control group converts at 4% on its own, the real lift is 2 percentage points, or 20 donors. The vendor caused a third of the result it is claiming. Without a control group, you cannot see that split, and you pay for all 60.

This is why the test design matters more than the headline number.

Questions to ask before you sign

Use this checklist in evaluation calls. Strong vendors answer plainly. Weak ones redirect to testimonials and benchmarks.

On measurement method

• Is the result from a randomized control group, or from total activity during the period?

• How was the control group selected, and was it comparable to the treated group at the start?

• What was the baseline conversion rate in the control group?

• Is the headline number gross conversions or net incremental lift?

On attribution

• How do you separate conversions you caused from conversions that would have happened anyway?

• How do you handle donors who also received our other appeals during the test?

• Over what window is a conversion counted, and does it include later cancellations?

On durability

• What is the retention rate of the recurring donors you convert at 6 and 12 months?

• Do you report revenue net of churn, or sign-ups at the point of conversion?

• Can you show lift sustained across more than one campaign or season?

On proof you can inspect

• Can you share a result from an organization like ours, by size, cause and region?

• Can we run a controlled pilot on our own file before committing?

• Will we get the raw data to verify the result ourselves?

What good evidence looks like

You want a claim you can repeat to your board and defend. Good evidence has four traits.

It is tested against a control group. It reports net lift, not gross activity. It accounts for churn, so you are measuring revenue that stays. And it comes from organizations close enough to yours to be relevant.

If a vendor can show tested lift on a file like yours and offer a controlled pilot on your own data, you are looking at a measurable claim. If the proof is a platform average and a few quotes, treat the number as marketing.

Run a pilot you can measure

The cleanest way to settle the question is to test on your own file. Split a representative segment into a treated group and a withheld control group. Run the vendor's flow on the treated group. Compare recurring conversions and retained revenue after a set window.

The result is specific to your donors, and it belongs to you. That is the number worth signing on.

The takeaway

Recurring-giving claims are only as good as the test behind them. Raw activity and platform benchmarks describe what happened somewhere; they do not prove what a vendor will add to your program. Ask for control-group-tested, churn-adjusted lift, insist on a pilot against your own file and buy the result you can measure and defend.