There are some assumptions about life and marketing that are comfortable and unquestioned: Other people share our values, the people we mail will be interested in what we have to say, advertising is usually based on truth.
Clearly we should question our assumptions, especially when facts appear to contradict them. When I started renting lists for mailings for a cataloger I used to work for I gathered several assumptions about the lists from the prior list manager at the firm. I believed them—until I started segmentation and modeling analyses for the customers already in the database, and some unarguable facts popped up.
Assumptions are often held about mailing and telemarketing lists. There is some consistency to a rental list—all the people on the file bought the same product, used the same service or donated to the same cause in a specific period of time.
Marketers typically have two main assumptions about lists. First, they assume that mailing a similar offer to a selection of the list will get a good response rate. But different offers for product, services and causes that are similar but not identical have a probability of getting a response, and you are risking a lot of money and effort on that assumption because you don’t know the probability.
Second, they assume that setting up and running a list test is time-consuming and expensive. They try to overcome the lack of real information by making more assumptions about rough similarities of the age, income and geography of the people on the list, which are really just thoughtful guesses.
You can improve the probability of getting a good response rate from your mailing by replacing your assumptions with real information.
First, extend your timeline to include time for testing and analysis. Second, mail a random sample of the list you like. You can check the distribution of your sample with the list owner’s or manager’s help to insure that your test will include equal proportions of the different age, income and geographic brackets. Third, mail your offer in a quantity large enough to get a significant response—50,000 pieces to get 500 respondents through a 1.0% response rate will give you enough actual responders to analyze. Finally, have your analysts create a simple model from those responders based on their demographics.
The model can be used by your list provider or service bureau to sort out the prospects most likely to respond from the ones that will rarely or never respond. The money you spend on the test and the modeling will be repaid both in the savings you’ll get from not wasting mail pieces on non-responders and in higher sales from the improved response rate you’ll get from mailing more to the prospects the model highlights for you.
Questioning your assumptions about your mailing lists can have several beneficial outcomes. You will learn about the makeup of the lists through the test mailing and analysis. You will have established a test, learn and model process that you can use over again with other list sources and with your own customers. And you will have replaced guesses with facts about your prospects that will save you money from wasted mail and make you money through better targeting of your mail to the prospects most likely to respond.
Bill Singleton is a manager of analytic services at The Allant Group, Naperville, IL.
