Maintaining A Mean, Clean List

Circulation Management, Nov, 1999

List hygiene isn't the sexiest aspect of direct marketing, but the significant savings realized over time through consistent cleaning and wise use of merge/purge can free up lots of dollars for creative marketing programs.

Memo

Here's a definitive list of the eight steps to a squeaky clean list, in order of importance, from Cary Zackman, director of sales for Experian (which offers some of these services):

1 Use National Change of Address (NCOA). Mailers who don't use NCOA are being penny-wise and dollar foolish, says Zackman, who points out that use of NCOA produces an average response lift of 400 percent. More than 23 licensees offer the NCOA service.

2 Identify more valid duplicates by using better software. In many mailings, duplicates represent between 1 and 4 percent of the main file. Software duplication identification capabilities vary widely. Make sure that the program you're using can identify these common, but frequently missed, types of dupes:

* First letter of the last name is different.

* Flip-flopped first and last names.

* Married/maiden name female dupes (same first name, different last name; same address), which account for between one-quarter of a percent to three-quarters of a percent of total dupes.

* Special situation dupes, such as two variations of a name within a zip code, one with box number, and one with a street address. (These account for about one-quarter to one-half of a percent of dupes.)

3 Identify more changes-of-address by using proprietary COA systems. With a proprietary system, an additional 1 to 2 percent of the address changes not identified by other sources may be attainable.

4 Use Nixie Elimination Service (NES) to identify close matches to the NCOA database and determine which are probable movers. These probable movers account for about 1 percent of a file, on average, and perform at only 30 percent of the response rate for a mailing as a whole.

5 Use Address Element Correction (AEC) in order to zip 4 more records. On average, each record that is zip 4 coded saves 5.8 cents in postage. Moreover, noncoded names perform at only 60 to 70 percent of the response rate for the mailing as a whole.

This can be done through U.S. Postal Service software or proprietary software. Each type of program can correct 30 to 50 percent of the non-coded records, and the two used in tandem can generally correct up to 50 to 70 percent of the non-coded records.

The USPS AEC identifies names totaling to about one-quarter to one-half of a percent of a file, compared with three-quarters to 1.5 percent for proprietary programs. And, because the USPS service is an intensive analytical process, Zackman recommends that it be used only on house files outside of the normal mail stream process.

6 Use multiple zip -4 engines. Another way to increase the number of zip- 4-coded names on a file, with the same benefits as those outlined above.

7 Use Delivery Sequence File (DSF) to identify questionable-delivery records. Non-matches to the DSF file account for, on average, between 2 and 6 percent of all records, and up to 40 percent of these do not get delivered. These records include those for vacant or seasonal addresses and those missing an important address component.

8 Use the Locatable Address Conversion System (LACS) to convert rural route addresses to street addresses. The average match rate here is about one-half a percent, and performance of these names is about 50 to 70 percent of overall mailing response.