What to Consider Before Finalizing Your 2011 Catalog Circulation Plan
Direct marketers love data, and direct marketing is a statistician's dream. Mailing millions of catalogs yields a rich pool of data to drive decision making. Direct marketers must learn how to use statistics to properly analyze data and to learn the pitfalls and shortcomings of that data.
Statistics are the foundation of circulation planning. Knowing the truths and fallacies of your statistics can show you how to plan circulation. Circulation planning is based on these simple statistical truths:
- Past results from mailings are the best predictor of future results. How a specific mailing list or list segment performed in the past gives an accurate prediction of how it will respond the next time you mail it.
- The more recent the previous mailing, the more reliable the data. That's why a record of how a list has performed over time is the basic metric for planning future circulation.
- Mail lists that responded above breakeven in the past; don't mail lists which responded below breakeven in the past. When you prospect for new customers at or above breakeven, try to find ways to make your profitable lists more profitable and mail to those rewarding lists more frequently. Mail deeper into your profitable lists so zero month to 12 month buyers respond above breakeven. Test mailing to those buyers older than 12 months, as well as mailing stronger offers to your more lucrative lists.
- Identify households that aren't responsive and suppress them. Catalogers rely on co-op databases to gather together all the transactions from thousands of catalogs. The databases can tell you the households that have stopped buying. If a household isn't buying from any other catalogs, it's not going to buy from yours. Suppressing nonresponsive households saves the cost of printing, paper and postage (roughly 50 cents to $1.00 for each household you don’t mail).
These statistics drive circulation planning. However, you can easily get tripped up in analyzing the vast sea of data that's available. Here are some things to consider when analyzing and testing catalog data:
What sample size do you need, and can you afford to test whether a list will yield a profitable response? Mailers typically test with a sample size of 5,000 or 10,000 mailing pieces. With response rates averaging 1 percent, a test of 5,000 pieces would yield 50 orders. If a piece costs 50 cents to print and mail, a test of 5,000 pieces costs $2,500. Tests with smaller sample sizes can yield fewer orders, so it's difficult to know how a bigger test of the same list will perform. List brokers will often suggest much larger tests of unproven mailing lists, but a test of 50,000 names costs $25,000. One of the keys to profitable mailing is learning as much information as cheaply as possible.