Donor acquisition: judging a channel by retained donors, consent and capacity

What to take away

Most acquisition reporting answers the wrong question. It tells you what a lead cost. It does not tell you whether that lead is still giving two years later, and that gap is where acquisition budgets quietly fail.

A channel that delivers cheap one-time gifts can cost more per retained donor than a slower, pricier route. The only way to see it is to follow each cohort forward and charge the channel for the donors it actually kept.

Set the ceiling before the campaign

Cost per retained donor is the number that decides whether a channel earns a second year. You can set an acceptable ceiling yourself with three figures you already hold.

Let L be the average lifetime value of a donor acquired through this channel. Let s be the share of that value the organization is willing to spend to acquire it. Your ceiling is L x s.

If L is 900 dollars and s is 0.25, the ceiling is 225 dollars per retained donor. A channel returning donors at 180 dollars passes. One at 310 dollars does not, however good its lead volume looks.

Run the same arithmetic per channel, not per campaign. A blended figure hides the channel that is losing money and protects it from the cut it has earned.

Where the money actually goes

Acquisition cost is rarely one line. Media spend is the visible part. The rest arrives as staff time, platform fees, payment processing and the thank-you and receipting work that follows every gift.

Two habits distort the total. Charging only media spend makes every channel look affordable. Charging the whole program team to acquisition makes every channel look unaffordable. Pick a rule, write it down, and apply it to every channel in the same quarter.

Consent, receipting and data residency

Consent, receipting and data residency are the three questions that stop campaigns after the creative is finished. Each has an authority behind it, and none of them is a marketing decision.

Under CASL, consent is express or implied, and implied consent runs on a clock that depends on how the address was obtained. PIPEDA and British Columbia's PIPA govern how you collect and store donor data. Quebec's Law 25 adds its own requirements for donors in that province, and Bill 96 affects French-language obligations.

Receipting sits with the CRA. Disbursement quota rules and receipt content are set there, not by your agency. None of this is advice you should take from a blog, including this one. Take it to a qualified Canadian charity lawyer or to the CRA before the campaign ships.

For donors across the border, the IRS sets the treatment, and a cross-border giving specialist is the right call. Do not let a US-facing appeal imply a deduction you have not confirmed.

Build the sign-off file

Every campaign should leave a file that answers the next campaign's questions without a meeting. Build it as the campaign runs, not afterwards.

The file is short by design. A page per campaign is enough if it holds the five items above.

Check the creative against the evidence

Program records decide what a public statement may claim. A single case story can illustrate one person's experience with permission. It cannot establish a general program effect, and a funds-raised total does not show what changed for participants.

Urgency needs a real deadline or a documented need. A countdown timer and an implied shortage manufacture pressure the program cannot honour.

The FTC guidance for marketers using reviews covers fake feedback, selective requests, conditioned incentives and paid ranking. Apply those checks before you name a vendor in a donor-facing page.

Match the response to capacity

The public message is the first step of a relationship another team has to serve. Map what happens after a person responds, and test the normal route plus one realistic failure before you increase reach.

Test the donation form against the W3C explanation of information and relationships. Visual structure must also be available to assistive technology, and that applies to the form, the gift array and the confirmation page.

When you buy the tools behind all this, the CISA software acquisition fact sheet sets out supply-chain and vulnerability questions to ask. The GOV.UK guidance on choosing technology adds ownership cost and data control. Neither approves a vendor. Both belong in the purchase record.

Read the dashboard without fooling yourself

Separate diagnostic metrics from decision metrics. Clicks, open rates and cost per lead tell you whether a page works. They do not tell you whether the campaign was worth running.

Decision metrics are cost per retained donor, thirteen-month retention by cohort, average gift in year two, and monthly giving churn. Attribution rules assign credit to a touchpoint. They are not proof that the touchpoint caused the result.

Choose an outcome window that fits the action. A same-day click diagnoses a page. Donor retention needs a year. If a channel fails the ceiling test, cut it even when its traffic rose.

Common questions

What is a good cost per retained donor?

There is no universal figure, and anyone quoting one without a source is guessing. Set your own ceiling as lifetime value times the share of value you will spend to acquire it, then hold every channel to it. Publish the current sector figures from your own program data or a named benchmark study.

How long before acquisition results mean anything?

Early indicators are diagnostic only. Retention, second-year giving and monthly giving churn need at least thirteen months of cohort data before they can carry a budget decision. Review at ninety days to fix the page, and at thirteen months to fix the channel.

What has to be settled before a campaign launches?

Consent basis, receipting treatment and data residency. CASL, PIPEDA, BC PIPA, Quebec Law 25 and Bill 96 all touch donor data, and the CRA sets receipting rules. Confirm each with a qualified Canadian charity lawyer or the CRA, and record the answer in the sign-off file.

When should a campaign be paused?

Pause when the public promise outruns service capacity, when a claim loses its program source, or when consent cannot be shown for a list. The named stop authority protects affected people first, then corrects the statement where it appeared, then documents the closure.