Welcome to Part 4 of our “Actually Useful Maths for Data Work” series.
Written by our very own Anthony Fawkes, this five-part series will provide fundraisers, data managers, analysts, and CRM heroes holding charity operations together with easy-to-apply solutions that strip away the stress from maths. The aim is to show you how everyday calculations (and a bit of logic) can level up your data confidence.
What Is Attrition Modelling?
It’s a way to understand how and when people stop engaging-whether that’s regular donors cancelling, event attendees not coming back, or email subscribers dropping off.
Where to Start: Look Back Before You Look Forward
Ask: When did they last give? How many months until they stopped responding? What’s the average donor lifespan? Do different groups behave differently?
Build a Retention Curve
Create a line chart for each donor cohort. Track what % of donors are still active each month. Compare different channels or campaigns to spot trends.
Key Metrics to Try
| Metric | What It Tells You | Why It Matters |
|---|---|---|
| Churn rate | % who stop giving each month | Helps predict drop-off and set goals |
| Median donor lifespan | How long supporters stick around | Target re-engagement before they leave |
| Attrition point | When most drop off | Perfect timing for interventions |
| Cost-adjusted value | Value after costs | Shows true profitability |
Included Costs
Track net value: include staff time, tech fees, campaign costs. Donor value isn’t just income- it’s income minus effort.
Let’s Talk Tools
| Task | Excel | Power BI | Python |
|---|---|---|---|
| Track last gift date | =MAXIFS() | Filter + DAX | pandas date functions |
| Retention curves | Manual cohorts + charts | Custom visuals | matplotlib cohort plot |
| Attrition rate | COUNTIFS, pivots | CALCULATE with filters | date diffs in pandas |
| LTV modelling | SUMIFS + cost fields | Net income per donor | regressions or calculations |
What To Do With What You Learn
Set up thank-you campaigns, create CRM alerts, improve messaging for high-attrition segments, and invest in channels with stronger retention.
TL;DR
Attrition modelling helps you anticipate supporter drop-off. Use it to act earlier, smarter, and with more context.
Coming Up Next in the Series…
- Post 5: Python For Fundraising Automation – Making Data Work While You Sleep
Want The Rest Of The Series Now?
The full series of posts is available to download right now! Sign up for our newsletter and we’ll send a copy straight to your inbox.
