Developing Business Rules for Matchback Allocation

Apr 28, 2007 3:14 AM  By

With the flood of web buyers arriving without source codes, matchback analysis has become the methodology for reading results. Matchbacks are simply matching the mail file of the circulation with the response file of orders received during the life of the catalog being measured. The key to accurate matchbacks is setting the business rules for allocating orders between catalogs and between catalogs and other customer contacts. The devil is in the details. Here are issues catalogers face in striving for accuracy in their matchbacks.

  1. Should you fractionally allocate orders when there have been multiple e-mail contacts as well as catalog contacts of the same customer? Rule of thumb is to look at the response rate of e-mails and see if allocating some of the orders to e-mail contacts makes a material difference. If e-mails to a buyer segment typically yield $.10/e-mail and catalog mailings to the same house segment yields $4.00/catalog, then fractional allocation doesn’t make any difference.
  2. How do you allocate orders when multiple catalogs are mailed in the same season? Problem occurs often for holiday fourth quarter mailings when the same house buyer segment might receive three catalogs spaced two weeks apart. If you allocate orders to the last catalog mailed, the catalog that is the last book in-home will get allocated too many orders, and the other two catalogs will only get two weeks worth of orders. A simple solution is to take all the orders for the holiday season and allocate them evenly across all the catalog contacts in that season.
  3. How do you measure the incremental sales from a catalog mailing? Always run a mail vs. no-mail test of house buyer segments to measure the incremental sales coming from a catalog mailing. Remember that some customers will order even if they don’t receive a catalog, and you need to know your underlying order volume you will receive even if you don’t mail a catalog.
  4. How do you measure to catalog end when the date range of your matchback ends before the end of the life of your catalog? Track source code responses by week and use that order curve to extend your matchback results to catalog end.
  5. Look at directional data to validate your matchback results and your allocations. Use promotional codes capture shopping cart responses as an alternative to catalog source codes. Look at original source codes to show the proportional response between different segments and to read results early in a mailing before matchback results are available. Look at your daily sales flash report to measure the increase in sales from a catalog drop. The daily sales flash is especially predictive of how well a mailing will do overall; a mailing that starts strong tends to stay strong and vice versa.

The basic matchback process is pretty simple – match the mail file and the response file. If you take care of the details of allocating orders, you’ll have the confidence in your matchback reporting to make the best business decisions about future circulation plans.

Jim Coogan is president of Santa Fe, NM-based consultancy Catalog Marketing Economics.