Which fundraising attribution model should your non-profit use?

Attribution decides which channels and appeals get credit when a donor gives. That credit shapes budgets, channel plans and how you defend spending to your board. Pick the wrong model and you will overfund the channel that closes the gift while starving the channels that started the relationship.

This guide compares the five attribution models non-profits use most, explains how to handle offline gifts from direct mail and events, and shows how to match a model to your programme mix and data maturity.

What is fundraising attribution?

Fundraising attribution is the method you use to assign credit for a gift across the touchpoints a donor saw before giving. Those touchpoints can include an email, a paid ad, a direct mail pack, an event or an organic search visit.

The model you choose sets the rule for how that credit is split. It does not change what donors did. It changes what your reports say caused them to give, and that is what your team funds next.

Which attribution model should a non-profit use?

Short answer: match the model to the decision you need the numbers to defend and to the data you can actually track.

• Use first-touch when your priority is acquisition and you want to see which channels bring new donors in.

• Use last-touch when your gifts follow a short, direct path and you need a simple, reliable baseline.

• Use position-based when both discovery and conversion matter, which fits most multichannel programmes.

• Use time-decay when donor journeys are long and recent touches clearly do the heavy lifting.

• Use custom multi-touch only when you have clean cross-channel tracking and the maturity to maintain it.

The sections below explain the trade-offs behind each.

First-touch attribution

First-touch gives 100% of the credit to the first interaction a donor had with you.

It answers one question well: where do new donors come from? That makes it useful for acquisition budgets and top-of-funnel channels like paid social, search and awareness campaigns.

The weakness is obvious. It ignores everything that happened between discovery and the gift, so it undervalues stewardship, retention and the appeal that actually asked for money.

Last-touch attribution

Last-touch gives 100% of the credit to the final interaction before the gift.

It is the most common default because most tools report it out of the box and it is easy to explain. For programmes with short paths, such as a single appeal email that drives a donation, it is often good enough.

The risk is that it overcredits closing channels and hides the work that built intent. If you judge every channel on last-touch, you will keep cutting the campaigns that create donors and keep the ones that harvest them.

Position-based (U-shaped) attribution

Position-based, often called U-shaped, splits credit between the first and last touches and shares the rest across the middle.

A typical split gives 40% to the first touch, 40% to the last touch and 20% to everything between. It rewards both the channel that found the donor and the channel that closed the gift, which suits most multichannel non-profits.

This is a strong default when you run acquisition and conversion programmes together and want one model that respects both.

Time-decay attribution

Time-decay gives more credit to touches closer to the gift and less to earlier ones.

It fits long consideration journeys, such as mid-value cultivation or legacy programmes, where recent contact tends to trigger the decision. It still recognises early touches, but it weights the finish line.

The trade-off is that it can undervalue genuine acquisition work when a donor takes months to convert. Use it when you believe recency truly reflects influence, not just because journeys are long.

Custom multi-touch attribution

Custom multi-touch assigns credit using weights or a model you define, often built from your own data.

Done well, it is the most accurate reflection of how your donors really move across channels. Done poorly, it is a black box no one trusts and no one can maintain.

Only adopt it when you have reliable cross-channel tracking, consistent identifiers and the analytics capacity to keep it current. If leaders cannot explain the model, they will not use it to make decisions.

How to handle offline gifts with matchback

Offline matchback is the process of connecting a gift that arrives offline, such as a mailed cheque or an event donation, back to the campaigns that likely influenced it.

Direct mail and events rarely carry a clean digital trail. Matchback closes the gap by comparing donor records against recent campaign audiences to assign credit those channels would otherwise lose.

Practical steps:

• Tag each mail drop and event audience with a clear code and date.

• Set a matchback window, often 30 to 90 days, and apply it consistently.

• Match returned gifts to the campaigns a donor was exposed to inside that window.

• Document your rules so the numbers are repeatable and easy to justify.

Without matchback, direct mail and events look weaker than they are, and multichannel donors get miscredited to whatever digital touch happened last.

Compare the models at a glance

Match the model to your programme and your decisions

Start with the decision, not the model. Ask what the numbers have to defend, then pick the model that answers it honestly.

• If you are defending acquisition spend, lead with first-touch and support it with matchback.

• If you run a broad multichannel programme, position-based is the safest default.

• If your gifts follow long cultivation, time-decay reflects reality more fairly.

• If direct mail or events carry your revenue, matchback is not optional, whatever base model you use.

Also match the model to your data. A precise model you cannot maintain is worse than a simple one you trust. Choose the most accurate model your tracking can actually support, then improve it as your data matures.

Practical takeaways

• Attribution is a budgeting decision, not a reporting formality.

• Last-touch is a fine baseline but a poor sole judge of channel value.

• Position-based fits most multichannel non-profits.

• Always run matchback if direct mail or events drive real revenue.

• Pick the most accurate model your data maturity can sustain, and make it explainable.

Conclusion

There is no single correct attribution model. The right one is the model that reflects how your donors give and gives your team credit numbers they can defend. Start with the decision you need to justify, match the model to your programme mix and data, and use matchback to keep offline channels honest. Get that right and attribution stops being a debate and starts guiding where your next pound should go.