When Response Drops—and Profit Rises

What?

At a large nonprofit mailing to 40k supporters, the team tested a prediction model to choose who should receive a house mailing. They optimized for lifetime value (LTV), not raw response. The test was statistically significant: response fell −7,66%, but average donation rose +17,86%. Revenue per letter increased +8,83%, and profit per letter grew +9,82%. The team accepted fewer gifts to earn more per letter overall.

House mailing optimization means deciding who to mail and when, using data and tests to raise money per letter.

LTV = lifetime value, the expected net revenue from a donor over time.

So What?

Postage and printing make mail costly; mailing everyone risks waste. A model that prioritizes bigger expected value per donor improves unit economics even when fewer people give.

Myth: prediction is for quick wins. Correction: it’s for confidence and flexibility—knowing who to approach at any time—making donor‑centered fundraising practical.

With added profit, it may be cheaper to reinvest in acquisition or reactivation to grow total active donors than to send more letters. Evaluate this with data, not instinct.

Now What?

This week:

1. Set success to LTV, and track four numbers: response, average donation, revenue/letter, profit/letter.
2. Run a model‑vs‑response‑rate A/B on a defined segment. Keep holdout controls. Test for statistical significance.
3. Decide reinvestment vs. more mail: compare cost per new active donor from acquisition/reactivation to the incremental profit per letter from expanding volume; choose the better unit outcome.

Results

• −7,66% response (representative test result)
• +17,86% average donation (representative test result)
• +8,83% revenue per letter (representative test result)
• +9,82% profit per letter (representative test result)
• Realized Potential: +xx.xxx€; Total Potential: +xx,xxx€ (as provided)
• Prediction model produced a statistically significant uplift

Key Takeaways

• Make donor‑centered fundraising your default.
• Optimize for LTV; judge by profit per letter.
• Use the 4R + Reinvest Test to decide.
• Response vs. LTV depends; models fit the strategic goal.

Curiosity Spark

How would your decision change if acquisition beats mailing on cost per new active donor?