Figuring out what data is available when marketing to high net worth individuals is the first step to putting together a successful targeted marketing program focused on this demographic. Most individuals with high net worth have gradually accumulated wealth over several decades. It may be due in part to a very good career, inherited wealth or wise choices in investments or a combination of all three. Through the years they have been marketed to by many different organizations and have made choices (most of them good) about how to handle their financial affairs.
Most high net worth or very high income individuals look to reduce their tax burden through tax advantaged investments such as real estate, IRAs, 401Ks, tax free bond funds, fixed or variable annuities and some may participate in limited partnership investments in movies, oil and gas, or other private placements. Beyond the IRA many may have significant holdings in their own 401K since contributions are not as limited. Over time they have become selective about what they respond to and based on our market research 90% or more are on the Do Not Call list.
However, they are not finished looking for new opportunities to preserve and increase their wealth. Targeting high net worth or very high income individuals who meet the criteria for accredited investors requires finesse because the simple techniques no longer work.
Basic List Selects for High Net Worth Individuals
Let’s start with the basic selects to get us in the “wheelhouse.” There are three basic types of selects available on a list; known, self-reported and modeled. Known data is sourced from public records such as deed information. Self-reported data is provided by the individual and may or may not be accurate; particularly self-reported age and income. Other known data to a lesser extent (due to the source as explained later) would be age and gender. A third type would be “self reported.”
Modeled Data for High Net Worth Individuals
Modeled data is inferred and many companies have proprietary models that infer many things about an individual. Some models use transactional data from credit cards; not only how much is purchased monthly but what kind of purchases they are; luxury goods versus low-end merchandise, what kind of store it comes from and the average individual purchase etc. Models infer wealth by deed information and some infer income producing assets. Other models use (in part) an extensive consumer survey data and compensate individuals for the information and then apply all of their attributes to find other similar individuals. Some models thousands bits of information on any household or individual. You probably notice how information is collected on you and the offers you receive from new companies that target your interests.
Who Uses Modeled Data to Target High Income, High Net Worth Individuals
Political offers can reach outside of their respective party’s house list of donors to select high income, high net worth individuals from lists (or selects on outside lists) that closely reflect their core base. They can test using a wide range of selects and then drill down on their better tests to find more key data. Colleges use market research to reach high net worth graduates, Private schools can market to high net worth potential donors or link religion and high net worth. Mailers who use surveys can increase response by mirroring the existing demographics of their house file or its better elements. Charitable offers can find more high-end, high net worth donors by overlaying these selects and mailing to their core donor’s relatives, neighbors or work associates and test each segment to determine their best results. Yes, one list source we use knows who your relatives, neighbors and work associates are.
Using Income Modeling to Infer Data
A good example of inferred data is income. No one truly knows how much income an individual makes besides that person and the IRS if they are honest. Many companies offer lists with an income select and what they use for a model varies widely and so does their output.
For example, we rented a list for a client from a well-known data compiler and one of the selects we used was an income of $400,000. We then had our service bureau use a second income model from their well-known provider. (More detail on getting the most out of a computer service bureau in a future post.) Basically a computer service bureau receives all the lists in a project and performs various processes such as a NCOA and other functions on all the lists in the project to refine the data.
The second model put some of the records in an income range of under $75,000 when the first source had them at over $400,000! Thus illustrating the difference between models and what data the model uses and how much importance they give each piece of information. For more on this interesting topic, please see our next post; Models, Models and More Models.