Risk-value segmentation takes into account the risk of losing a donor and the potential value of that donor.
World Vision developed and used this segmentation model to get better return on investment from several direct mail campaigns. The results included average gifts up to $30 higher and response rates 10-20% higher.
Thanks to supporter growth manager Bernadette Kennedy for sharing on this at the recent F&P Forum. I thought this model could be useful to other charities, so I decided to follow up on her generous offer to speak to their Marketing Intelligence team about it. Senior marketing intelligence manager Kate Penglase was kind enough to answer my questions about it.
Basically, the model uses two components.
1. Calculating risk
Because 60% of World Vision’s supporters are child sponsors, the model takes into account the risk of child sponsors cancelling if they receive an additional appeal.
World Vision already had a lot of data relating to cancellation risk factors including:
- Falling behind in monthly donations
- Age – older sponsors who may be pensioners or younger sponsors who are less stable
- Sponsors who unfortunately experience a high number of children dropping out of the program.
Risk scores are assigned to each variable. Based on the score, a sponsor with a combination of risk factors may be excluded from extra mailings.
2. Calculating potential value
This was more complex. The traditional RFV (recency, frequency and value) model looks at the actual value of donors’ gifts.
However, in this model, the idea is to look at a donor’s potential value. This is done by looking at giving behaviour in two 18-month blocks and benchmarking it back to the average.
If the donor is under the average, he or she has not yet reached their full potential value. If the donor is above the average, he or she has “maxed out” their potential value and should be excluded from further appeals.
So does that mean you can ask too much?
One of my biggest questions around calculating potential value was it appears to challenge the traditional fundraising notion that the more you ask, the more you get.
This model has an in-built assumption that donors have finite resources. That there may be a point where a charity is actually asking too much. (Many fundraising consultants would be horrified at this).
What was interesting to me was the contrast in response rates. In one appeal, World Vision identified 70,000 donors who had supposedly maxed out their potential to give.
Because 70,000 donors were too many to remove from a campaign, part of the group was retained.
The group under their potential value had a response rate of 23%. The maxed out value group? 15%.
So what does this mean? Well, World Vision says it means the model needs tweaking. But with all other things being equal, it appears to mean that the more you ask, the more you do get. But the extra you get does start to decrease the closer you get to a donor’s potential value.
How to apply the risk-value model
World Vision uses it in two ways:
- To identify low risk, high value supporters for mailing premium packs
- To identify supporters over their potential value to exclude from extra mailings.
In the latter case especially, it’s not necessarily about never mailing these donors with extra asks. It’s about identifying those who are most likely to respond. That means you save money on mailing donors and get a better return on investment.
What do you think? Have you ever tried a similar segmentation model? What were the results?