Key points
- Next-best action is the recommended next step for a donor, segment or campaign, given what the data shows today.
- It has four parts: prioritize, prescribe, orchestrate and learn.
- It turns a one-time segmentation into a decision that is made again as the facts change.
A working definition
Next-best action is the recommended next step for a donor, a segment or a campaign, chosen from everything the data shows today. It answers one practical question: what should we do next to raise the most while respecting the donor relationship?
The answer might be a message, a channel, an offer, a time or an ask amount. It might also be a decision to do nothing: to leave a donor out of a mailing they are unlikely to answer.
Why a score is not enough
Many teams already hold more data than they can use. They know who gave, who lapsed and which channels performed. What they lack is a reliable way to turn that into the next decision.
A fundraiser on a deadline does not need a probability by itself. They need direction: send this appeal, pause this cadence, call this donor, suppress this list, test this channel. When AI stops at a score or a dashboard, it hands a stretched team one more thing to interpret. When it recommends the action, it works at the speed of fundraising.
The four parts
Prioritize
Decide which donors or segments matter most now, by likelihood, value, timing or risk.
Prescribe
Recommend what to do: the message, channel, offer, cadence, ask, suppression or hand-off to a person.
Orchestrate
Move the recommendation into real work across mail, email, social, SMS, major gifts and reactivation.
Learn
Measure what happened and feed the result into the next round.
The fourth part is what separates next-best action from a one-time segmentation. Donor behavior shifts. So do channel performance, budgets and the economy. The recommendation is made again as the facts change.
What it looks like by program
- Annual giving: which donors are most likely to renew, upgrade or convert to monthly giving.
- Acquisition: which prospects deserve the investment and which to suppress.
- Reactivation: when to re-engage, which offer to use, and when to stop mailing chronic non-responders.
- Leadership: one decision framework, so fundraising, communications and operations act from the same intelligence.
The gain is more than efficiency. Teams that work from the same recommendations stop pulling in different directions.
How to start
- Pick one program and one decision, such as who to include in the next reactivation mailing.
- Build a model on your own past outcomes for that decision.
- Turn the scores into a plain rule the team can follow.
- Hold out a control group that gets the usual treatment.
- Compare net revenue between the two groups, then repeat with what you learned.
Squark returns plain-language, do-this-next recommendations with each score, and your team approves what goes out. The use cases show the decisions teams start with.