Who Are They?
July 15, 2008 Don Austin
At some point, most nonprofits ask the question, “Who are my donors?” It seems intuitive that if you know the characteristics of your donors you can market to them more successfully.
Answering this question usually means, “profiling” your donors. While this might sound easy, the process is not always straightforward. Profiling involves, first, overlaying demographic and lifestyle data on your donor file. Second, in the profiling step, you will have to choose between two methods to develop a picture, or pictures, of your donors.
Before you decide to begin this process you should ask yourself how you will specifically use the information and how you will justify the cost. You might find that a simple overlay of donor age will suit your needs.
There are a handful of companies that can append demographic and lifestyle information to your donor records. Frequently these data are shared among these companies, so, for the purposes of profiling it is difficult to say there is a best source for the information. It is more important to understand the general quality of the data that are sold. The data for overlays come from three sources:
- Self-reported information;
- Data from public records, and;
- Census data.
Self-reported information can come from product registration cards that consumers send in after purchasing products or from large-scale surveys to which consumers respond. These are frequently the source of lifestyle information.
You might recall filling out product registration cards yourself. There could be questions asking about your interests and hobbies as well as age, income, family size, etc.
Data from public records can include age, from state departments of motor vehicles, and information from real estate transactions, for example.
The third source, and least accurate, is Census data. The Census is conducted every 10 years, so the data age as time passes. In addition, the vast majority of households are asked to complete the short Census form, which does not collect very much household information. Only a small percent of households complete the long form. This form asks very detailed information about all of the family members present in a household and about the dwelling itself.
Data from these sources, millions and millions of records, are combined at the household level. Since there will be data absent from most households, information like income or home value must often be estimated or extrapolated from census data. Clearly, the information available from each American household is not necessarily accurate, nor consistent.
When an organization sends donor names and addresses to a company for a data append the percent of households that match the company’s database is usually between 80 percent and 90 percent. However, that does not mean that all of the data elements in the database are present for each household.
Age is one of the variables that have high availability. Of the donor names provided, between 70 percent and 90 percent will have a match for age of head of household. Gender is also present on most overlaid records because it can be deduced from the first name. Estimated income has a high match rate because it is often inferred from nearby households with a known income or from census data. Also, some companies have proprietary models to estimate income from other, more readily available, data.
Known marital status might be present on about 50 percent of the donor records while for another 40 percent the marital status is inferred. Occupation could be known for between 20 percent and 30 percent of your donors and educational level for 20 percent or less. Lifestyle information, such as an interest in golf, pets, travel, etc., is likely to be present for just a small percentage of your donors.
When you select the data elements you want appended to your donor records, you should select those that are available on the highest percent of households and can be useful in marketing.
To summarize, the information that you receive about your donors will be spotty in most cases. To gauge the accuracy of the data, it is a good idea to include the names and addresses of staff members in the overlay. Then you can compare actual information to what appears in the overlay.
There is another decision to be made after data have been appended to your donor file. What method will you use to characterize your donors? Two options are profiling and clustering. You will probably have to send your data to an outside company for analysis. Please respect your donors by not including names and addresses if you do this.
In the profiling method the data will be described one variable at a time. For example, 70 percent of your donors are older than the age of 65; 65 percent of your donors are female; 30 percent have some college education. These data can be presented as bar charts and even compared to the U.S. general population so you have an idea of how different your donors are to the U.S. as a whole.
This level of information has somewhat limited marketing usefulness, however. You won’t know how many donors are married, older than 65, and have incomes of more than $70,000, for example. This might be an important donor segment for some organizations. Many organizations have a segment of donors that is made up of affluent families with children, who make larger-than-average contributions. This would not be evident from a donor profile.
Clustering is a statistical method that groups like donors into clusters or segments. It would tell you if you have a segment of donors that is made up of affluent families with children, who make larger- than-average contributions and what percent of your donors are in this segment. It would tell you what percent of your donors are single females older than 60 with no children and low incomes. It is useful to keep the numbers of segments that are developed to no more than five or six.
Most organizations would not want to develop more than this number of strategies for their donor base.
The information derived from cluster analysis is more actionable that that gained from profiling. For example, organizations can go out to more affluent families with more personalized messages and appeals. Often the characteristics of a segment can be used to select prospect names from a compiled database.
With either method you will have gone a long way toward answering the question, “Who are my donors?” But, ultimately you will come back to the question of how to specifically use the information to improve your fundraising. Like other marketing analytics, profiling and clustering are just tools, not ends in and of themselves. NPT
Don Austin, Ph.D., is vice president of Client Strategy at May Development Services in Greenwich, Conn., a division of Direct Media. His email is firstname.lastname@example.org