Tales From the Crypt

For those of us who manage prospecting campaigns, there is a span of a month or two between ordering lists and dropping the direct-mail appeals in the mail stream. Usually that time is spent putting the finishing touches on the creative packages and getting the materials ready to go.

But while we fuss over the creative, something else is happening that’s every bit as important to the campaign. Down in the data crypt, thousands, and perhaps millions, of names from many sources are brought together for a complex process called a merge/purge.

The goal of the merge/purge is to make sure that each prospect gets only one piece of mail and that current donors don’t get any at all.

It sounds simple. After all, we have fast computers and great software that make quick work of combining dozens of lists, weeding out the duplicates and producing the output we need.

The fact is that every merge/purge is unique. Each reveals data about the quality of the process, predicts the outcome of the campaign and gives clues that help smart direct marketers plan for the future.

Quality control comes first
“If you can only afford to do one thing, make sure you NCOA [National Change of Address] all files before you put them into a merge/purge,” advises Andrew Mednick, managing director of Automated Resources Group, a database-marketing services

company in Montvale, N.J. “At minimum, all files should be run through CASS-certified address- standardization procedures to allow the merge/purge software the best chance of finding matches on names that have variations between multiple input files.

“You’ll likely improve deliverability and gain postal discounts by providing valid ZIP codes, ZIP+4 and other data which allow for mail automation,” he adds.

Mednick warns mailers not to ignore what industry pros used to call “NCOA nixies” — now referred to as “possible movers.”

“These are names and addresses from the donor file that come close to matching to the NCOA database but exhibit one or more differences in the comparison,” he explains.

The U.S. Postal Service cannot return a new address for a “nixie” because of the uncertainty surrounding the match, no matter how slight.

But some cagey mailers try sending to nixies on the off chance that a keying error occurred. (It’s possible for a donor to still live at the original address even though she showed up as a possible mover. If you do mail nixies, make sure to give them a unique source code and measure the performance of the test.)

Mednick adds that a number of other useful file-hygiene services are available depending on the specific needs, such as mining for deceased names. Check with your merge/purge provider to discuss extra services that might be of value to your specific campaign.

And don’t forget to suppress the Direct Marketing Association’s Mail Preference file from the output of the merge/purge. You’ll eliminate the names of consumers who’ve told the DMA that they don’t wish to receive fundraising appeals. You’ll lower your costs, be better stewards of your resources and lessen the flack that can come from mailing large acquisition appeals.

How clean is clean?
Sophisticated mailers already know that a “clean” file with accurate names and addresses means lower costs and higher revenue per piece mailed.

But what are the practical limits of a clean file? When the NCOA report comes back with your donor file, what is a good passing grade?

Some consultants will argue that no more than 3 percent to 4 percent of a file should need correction at any point in time. However, Mednick argues that a file that comes back after an NCOA cleaning with changes in a “single-digits percentage” was already pretty good.

Searching for clues
Big, bulky merge/purge output reports contain a wealth of information that helps list-media planners and brokers evaluate a campaign before it goes in the mail.

“I immediately look for lists that show a high percentage of single-buyers,” says Amy Houke, director of list and print media services for DMW Worldwide. “Prospect lists that contain a large percentage of names that don’t also appear on other prospect lists could mean the list will not perform well. It could mean the list is not up to date. More likely, the list contains names of people who simply aren’t good prospects.”

According to Houke, it’s a double-edged sword: “We want to find new prospects, but it’s hard to find large quantities of new donors among people who don’t already give to many organizations.”

Usually the best lists have the highest percentage of matches against your own donor database and prospect files.

But before your next prospect mailing, rank your prospect files by duplication rate against the others. Those that have the highest dupe rates will be your star performers. Remember, though, high duplication rates mean you must rent and exchange more names to get the desired quantity in the mail.

Houke recommends looking carefully at certain lists that dupe hard against one another.

“You may want to order just one list among a group that tends to have extremely high duplication against one another,” Houke says. “Unless you can secure an agreement from list owners to only pay for the net names you use, it probably makes sense not to order lists that have extremely high dupe rates.”

Don’t forget to suppress
Suppression files are lists of people who you don’t want to get your prospect mailing. Besides the DMA Mail Preference file, you’ll also want to suppress your donor file.

But be careful when selecting names: Only select current donor names and lapsed donors who you intend to solicit during the campaign. Don’t suppress long-lapsed donors unless you plan to mail them, because they could come back to you through a prospect list.

Tom Hurley is president of the not-for-profit division of DMW, a full-service, direct-response advertising agency with offices in Wayne, Pa.; Plymouth, Mass.; and St. Louis. You can reach him at 774.773.1200; or via e-mail at thurley@dmwdirect.com.