Segmentation of your database – The key to personalized fundraising

When ensuring that your database is well maintained, it is worthwhile to delve more deeply into the donor data. A useful and common approach is segmentation of your database. The idea of segmenting the database can generate enormous profits for those charities that do it properly. People are all different, even though they may support the same charity.

Segmentation could be defined as: “searching for groups of donors who resemble each other as much as possible within the group itself and differ as much as possible from other groups.”

In the context of the database, a segment is a group of similar donors distinguished from others because they are different in some way. Treating all these different people exactly the same does not make much sense. Effective segmentation is therefore the key to making fundraising work. It allows you to develop a different fundraising approach to each segment. In an economy where many non-profit organizations are competing for the same donor pool, segmentation enables you to personalize and prioritize your fundraising actions in order to maximize total giving. Moreover, the same model that segments your donor file can be used for donor acquisition. This will enable you to find other potential contacts that are similar to your best current donors.

Market segmentation in general

According to the principles of marketing as defined by Philip Kotler and others, non-profit organizations should consider each donor as a potential unique (mini) market. Each donor has its own unique needs and desires. In an ideal world, a charity should design separate fundraising programs for each donor. However, such complete market segmentation is not very realistic with regard to efficiency and effectiveness. Non-profits should accordingly define larger groups of donors who distinguish themselves by their needs and responses.

There are countless ways to segment a market and to segment a donor database. Kotler defines the following primary variables for segmenting consumer markets:

1. Geographic variables (region, district, city, rural versus global, etc.)
2. Demographic variables (age, gender, life cycle, income, education, etc.)
3. Psychographic variables (social class, lifestyle, personality, etc.)
4. Behavioral variables (consumer/donor behavior, claims, loyalty, degree of interest, etc.)

Many more variants and combinations of variants could be possible.

Segmentation criteria in fundraising

In fundraising, different segmentation criteria are in use. They should not necessarily be mutually exclusive. What are the most common and what trends can be expected for the future?

1. Demographic variables
Segmentation of your donor database almost always starts with demographic features such as age, gender, socioeconomic class, family size, income, education and occupation. Because these demographic data have been collected over many years, a great deal is known about the giving behavior of each group and their likely needs and preferences. All this information can be included in your database. Age tends to be a reliable indicator of the sources of information people use, the communication channels they are most comfortable with and the social influences they are susceptible to.

2. Behavior
It is also possible to segment by donor behavior: for example on how people performed in the past, by their most recent gift, by the size of their last gift, by how often they have given and by what they have given. This way of clustering is seen at the RFM segmentation and Donor Pyramid (see below).

3. Lifestyle and profiled data
Interest and personal attributes such as lifestyle and data derived from analyzing the lifestyle could also be an efficient way to cluster donors together. This form of data enrichment could e.g. reveal which donors are wealthy or which donors childless.

4. Predictive modeling
Segmentation criteria are changing constantly, but the solid base remains. We are currently heading towards predictive modeling, in which donor groups are clustered on the principle of expected behavior.

5. A future trend
The digital revolution and new media have resulted in a split in the (donor) market. Instead of push communication, we are increasingly heading towards what is known as pull communication, in which donors are looking and asking for information themselves rather than waiting until the information is provided unexpectedly. As a consequence, segmentation of the donor database may well become useless. A non-profit organization should instead focus on providing all necessary information on demand via various channels. In other words: focus on messaging tailored to the donors.

RFM

A solid and popular way of segmenting a donor database based on donor behavior is what is known as RFM segmentation. RFM stands for recency, frequency and monetary value and originated as a system of awarding points to evaluate mail-order customers in the USA.

RFM monitoring assigns a standard value to every donor on your list and then to watch their progress as your organization tries various methods of upgrading. The scores obtained by this method make it possible to predict future donor behavior. Many direct marketers see RFM as the basis of non-profit database segmentation, extended to include additional segmentation variables.

The Donor Pyramid

Another familiar way to classify donors is the Donor Pyramid. The Donor Pyramid shows the relationship between the number of donors and the size of donations, the different donor groups and the best method of donor care. This method of segmentation is based on the RFM technique and offers various fundraising strategies for each donor segment.

The Donor Pyramid is useful as a simple way of illustrating the traditional idea of the donor life cycle, the fact that a donor’s involvement starts with a general query and then moves or is led through various stages of ever-increasing involvement to the ultimate gift. Of course the pyramid is a simplification.

According to Ken Burnett in his book Relationship Fundraising, the pyramid should be seen as a ladder of involvement in which a fundraiser can move the donor up. In contrast to the shape of a pyramid, supporters are not necessarily being lost along the way. But of course we think it’s unrealistic to say that all donors will be upgraded to the top.

Life Time Value (LTV)

Segmentation based on LTV is a standard for defining the financial value of the donors. It shows the difference between future income from the donor and the amount invested in the donor (i.e. the ROI). Where RFM and Donor Pyramid segmentation look at the past, the LTV segmentation looks at the future. How much could a donor (group) bring in, if investment started today? Of course the future value should be calculated based on the present value.

Wealth Overlay and major donor segmentation

Your donors definitely include many wealthy people who are not necessarily already donating substantial amounts. In 2006, SAZ developed a new way to segment your database together with Elite Research. The so called Wealth Overlay filters your very wealthy donors from your database.

The result of this overlay is a group of existing donors with the financial capacity to make large donations, provided that you approach them properly. The Wealth Overlay is therefore the first step in major donor fundraising in order to bring these donors to the top of the Donor Pyramid.

SMART Segmentation

Once you have decided to start segmenting your database and determined which criteria or method to use, turn your attention to the information provided by your database. The segments you define should be SMART – specific, measurable, agreed upon, realistic and time-based. In brief this means that your segments should be relevant – otherwise segmentation is completely useless.

First of all, a segment is relevant if it is sufficiently different from the other segments. In addition, the segments should be homogeneous, substantial, stable and attractive. And last but definitely not least, segments should be accessible and quantifiable. Once the segments are formalized, a fundraiser should choose several of them as target audiences for the organization’s next fundraising campaigns.

Sources: Principles of Marketing by Philip Kotler, Gary Armstrong, John Saunders & Veronique Wong, Relationship Fundraising by Ken Burnett, Building Donor Loyalty by Adrian Sargeant and Elaine Jay, Partnership Fundraising, fundraising with a system by SAZ Dialog AG Europe.

This is the sixth post in our series Modern Database Management – provided to us by SAZ Marketing. The previous articles can be found here.