The president of Woodworker’s Supply, John D. Wirth Jr., Ph.D., has reason to be skeptical of software programs that label addresses “undeliverable.” In a test conducted during fall 2001, the Casper, WY-based tools mailer found that response from some segments of so-called undeliverables exceeded response from the house file’s deliverable names.
How can that be? It seems that undeliverability, like beauty, is in the eye of the beholder. More to the point, depending on the list hygiene and screening criteria used, undeliverable addresses often end up being valid.
What’s more, “a lot of these allegedly undeliverable addresses respond well because the customers receive so little mail,” Wirth says. “The competition isn’t mailing them, because they’re flagged as undeliverable or questionable.”
When processing mailers’ house files and prospecting files for undeliverable addresses against U.S. Postal Service-licensed programs such as the National Change of Address (NCOA), Delivery Sequence File (DSF), and Address Element Correction (AEC), some computer service bureaus simply flag address records as either “deliverable” or “undeliverable.”
Others go into more detail, however. Many NCOA users take advantage of the nixie option, in which codes indicate the reasons some records didn’t meet the strict USPS matching guidelines. A nixie code could indicate that a record would have matched except for a minor typo in the surname — “Smitth” instead of “Smith,” for instance. In the case of such discrepancies, the mailer may well decide that the address can easily be corrected and mailed.
Likewise, some users of DSF, the Postal Service’s database of all known mailable addresses in the U.S., rank results by varying degrees of potential deliverability. Marshfield, WI-based service bureau Donnelley Marketing Catalog Vision, for instance, uses a three-point scoring system, deeming addresses “deliverable,” “questionable,” or “undeliverable,” says vice president of analytics and marketing Keith Pietsch.
Farmingdale, NY-based service provider Anchor Computer ranks records that have been run against DSF on a scale of one to five. As sales executive Sari Schenker explains, a “one” is an address deemed “extremely deliverable,” in which a consumer’s name, street address, city, state, and zip + 4 code all match perfectly with the DSF. The higher the number, the more elements of the address — such as zip + 4 code, street address, or homeowner’s name — don’t match. “If you give me a file of 1 million records, I might find 30,000-50,000 records that are threes, fours, or fives,” she says.
Testing — 1,2,3, testing
Schenker recommends that her catalog clients test-mail to 50%-80% of those records that come up as a three, to 50% of those that come up as fours, and to 30% of those that are flagged a five. “We never take a position that a record is undeliverable,” she says. “It could be, for example, a brand-new housing development that hasn’t hit the post office’s file yet.”
Woodworker’s Supply also advocates testing. The cataloger, which has its house and prospecting files processed through the DSF at the time of the merge/purge, found 90,000 out of 2 million records flagged as “questionable” or “undeliverable” during its fall 2001 mailing. To see which of the 90,000 names could, in fact, be mailed, Woodworker’s test-mailed 26,000 of the undeliverable names and compared the results with those of a control group of 26,000 “deliverable” names from its house file.
The test showed that in some segments, particularly the less recent names, response from the “undeliverables” exceeded that of the company’s deliverables, making it worthwhile to continue mailing to questionable names. Wirth plans to conduct another test later this year.
“To get a good sample, you need a big mailing because you’re dealing with just a proportion of the merge/purge output,” Wirth says. “If you’re only mailing 100,000 or 200,000 books, you can’t get viable results because the sample size is too small.”
Costlier to mail
Not all catalogers believe in the viability of mailing to questionable or undeliverable addresses. One such mailer is Martin McClanan, former president of gifts cataloger RedEnvelope and now executive officer of San Francisco-based art materials and gifts mailer Flax Art & Design. McClanan is skeptical of such addresses, because “if you can’t barcode them, your postal costs would be much higher,” he explains. “And that makes the responsiveness difficult to justify on a return on investment basis unless they were extremely responsive.”
You may get a couple of good prospects out of mailing to undeliverable addresses or not, McClanan concedes, “but for the most part you’ll be mailing at first-class rates because you won’t be able to get the [technically undeliverable] addresses through a postal sort.” That could translate to paying $0.78 to mail a book as opposed to $0.20-$0.25 for standard mail, he says, “and it doesn’t pay to do it.”
Woodworker’s Supply’s success with mailing names that showed up in the DSF has been “exceptional,” notes Schenker of Anchor Computer. But other mailers might choose to complete their use of DSF with AEC, which tries to correct the erroneous or nonmatching elements of “undeliverable” addresses. That way, they could fix the addresses and then include them in bulk catalog mailings.
Of course, using AEC increases the premailing costs, Schenker adds. You would need to calculate, based on the number of records and other factors, whether using AEC would be more cost-effective than mailing the catalogs in question at the first class rate.
For the most up-to-date industry news, posted every weekday, visit www.CatalogAgemag.com.
To have industry news delivered to your desktop every week, subscribe to the free Catalog Age Weekly e-newsletter at www.CatalogAgemag.com.