Sometimes what you take out of a list can have a more significant effect on response than what you add to it. After all, who wants to waste valuable marketing dollars on names that are not likely to respond to your offer?
While you can probably guess some of the obvious groups of people whom you would not want to mail, you might be surprised at what other response traps could be lurking in a file.
With the help of a few list/database experts, I've compiled a collection of suppression criteria for direct mail lists. The first five ideas are applicable to all direct marketers, while the remainder are strategies that are specific to a company's list usage history, audience base or product category.
Suppress: Do Not Mail Requests
Of course you are already using external and internal pander files. But just in case you aren't, let's review this best practice.
In the course of your direct mail activities, you probably have received requests from prospects and customers to be dropped from your mailing list. These names should be assembled in a file that you run against all future prospecting and retention efforts.
In addition to your own pander file, the Direct Marketing Association (DMA) maintains a file of names and addresses of U.S. citizens who have asked that companies no longer send them direct mail campaigns. Called the Mail Preference Service, this file lists individual requester's names for five years.
While use of the file is a requirement of all member companies, they do not have to suppress their own house file names, says Jim Wheaton, principal of Wheaton Group and cofounder of Data University, which offers on-site database marketing training, webinars and consulting.
Wheaton explains that you own your customer and inquiry names, so you do not need to run them against the DMA's MPS file. But be certain to bump any prospecting files against this external suppression list.
Suppress: Prisons, APOs/FPOs, Deceased
Need I point out the obvious? Deceased people cannot respond to your offer, thus this is a waste of a direct mail package and a hurtful reminder of the loss of a loved one
Other no-brainer suppression segments are APOs/FPOs, which are military addresses, and prisoner addresses.
Paul Goldberg, president of direct marketing consultancy P-J Promotions, points out that rental lists should already be cleaned of these names, but don't take it for granted. When dealing with a new list manager or owner, be sure to ask if these names have been removed.
To keep your house file in tip-top shape, you can go to a computer service bureau to have prison addresses, military base addresses and names of the deceased purged from your file.
Suppress: Active Customers
If you're looking for new business, you will want to suppress your current customer file, notes Goldberg.
You might even expand that criteria to include recently lapsed customers, says Jean Tremaine, vice president of strategic development, Response Media Products. Since their activity with you just died down, these names might be more appropriate for welcome back offers than for messaging that treats them like new prospects.
Conversely, Tremaine suggests, many marketers have found precisely the opposite strategy to work better, since some people respond best to sweeter acquisition deals.
Suppress: Undeliverables
Just as there's no point in mailing to deceased people, why leave records on your file that are undeliverable as addressed?
"When assessing your house file, especially those customers who have not purchased recently, keep in mind that some might not be responding because they've moved and your efforts aren't following them to their new address," says Wheaton.
"For this reason, if you are using external data to improve your insight into your house file, you might want to note which older-recency names do not match," says Wheaton. "This is especially true if the external data has recently been double-source verified. A non-match increases the probability an older house file name has moved."
One way to eliminate records that are not deliverable is to run your list against a proprietary nixie file. While these files don't provide a forwarding address, Wheaton explains, they can help you suppress bad addresses before you mail.
Yet another deliverability factor to assess is records that are categorized as multi-family dwelling units but that don't have an apartment number. In such instances, you know there's a high probability that the mailing won't reach the intended recipient, says Wheaton.
Your computer service bureau should be able to tag such records, so you can drop them off your prospecting efforts.
When it comes to rented lists, Tremaine advises that all list orders should note that you will not pay for records that do not survive list hygiene processes. These reasons may include bad ZIP codes, data entry errors, incomplete addresses and intra-file duplicates.
Suppress: Bad Debt Risks
The level of financial risk you're willing to assume is particular to your company alone. Goldberg observes that some direct marketers keep a file of people who have not paid for products or services they ordered, either in full or in part.
It is also possible to order lists of people who have bounced checks.
Tremaine points out that some of her financial clients do pre-screens for offers. She explains that these marketers always negotiate up front a net-name deal for the percentage of prospects who won't meet the screening criteria "that way they don't pay for names that aren't profitable to mail.
Suppress: Business or Residential Addresses
A more particular type of suppression is that of eliminating business addresses on consumer mailings. One way to accomplish this, says Goldberg, is to flag all four-line addresses.
Since this won't catch some of the businesses that are home-based, Wheaton explains that marketers can rent compiled files of small office-home office (SOHO) businesses to bump against their house files.
A further consideration, he adds, is that some business people order gifts for their clients once a year. While this business is not frequent, it can be a significant profit stream for consumer marketers. Once you know which records are seasonal business, you can decide when it's appropriate to suppress these names from your more regular mail drops.
At the same time, business-to-business marketers often are not interested in mailing to residences, or at least not contacting them as frequently, even if these addresses belong to SOHOs, explains Wheaton. Not all home-based businesses offer the same long-term profitability to a b-lo-b company, so they are an option for suppression.
Suppress: Uniques
"A basic rule of direct marketing is that any name that shows up on two or more rental lists can be considered a multi-buyer. A name that shows up only once is considered a unique name," says Wheaton.
Generally, he continues, multi-buyers respond one and a half to two times higher than uniques. Some direct marketers will eliminate uniques on rental lists that are marginal, he explains, and especially if they can get a net-net list deal.
Companies also can model their house files to determine which customers are not likely to respond based on historical purchasing data. Those segments that score the lowest could be matched against a rental file or cooperative database, Wheaton explains, and the hits might be upgraded as selections to leave in the mail drop.
The strength of this approach depends on the recency of the house file name. But making a decision on a less recent customer based solely on your in-house data may not present you with the full picture of that person's offer-worthiness.
Suppress: Male or Female
If you have a product that skews heavily to women or men, then you can suppress either gender based on first names.
Again, Wheaton notes that major service bureaus maintain large first-name look-up tables, where names like Jim are tagged as male; names like Suzy are tagged as female; and names such as Pat are marked as unknown or either.
For marginal lists, this suppression approach can work very well in making the lists payout.
Suppress: Previously Mailed Files
Depending on how often you mail, and how often you mail the same lists, you might want to suppress names from your previous mail drops, says Tremaine. For example, if you're using the same list in January that you mailed to in November and December, it would be worthwhile to bump out those names previously rented.
Suppress: Rural or Urban
Quite simply, if you know that a significant portion of your customer base is rural, then it might be a good idea to suppress names from major metropolitan areas, offers Wheaton. The same concept applies in reverse, but both are dependent on your product or service.
Wheaton notes that you can find companies that turn census bureau data into carrier routes to identify market segments with particular demographics. (See "What Census Bureau Data Can Do for You" on p. 20.) Of course, whether you want to suppress or select these areas is up to what you know about your product and your customer base.
What Census Bureau: Data Can Do for You
Once a decade, the U.S. Census Bureau collects a wealth of information on the U.S population. Marketers can choose from 350 variables that allow them to gain a better idea of where and how their customers live, says Jim Wheaton, principal of Wheaton Group.
To give you a sense of the depth of census data available, here is just a sampling of the variables:
- percentage Dutch ancestry.
- percentage Swedish ancestry.
- percentage adults separated/divorced.
- percentage adults 25+ w/some high school (no diploma).
- percentage persons enrolled in private schools (age 3+).
- percentage adults in active military service.
- percentage households on social security.
- percentage seasonal/recreational vacant units.
When Not to Suppress
It's pretty obvious that the more you suppress, the fewer names you have to mail. And there are times when it doesn't make sense to remove names from a file.
For example, says Goldberg, your best-performing lists that you mail to repeatedly have shown an affinity for your product that requires no extra whittling to get the desired lift. You don't want to remove names that are likely to respond.
Tremaine points out that it's not the best idea to heavily slice and dice untested lists before you find out if they work for your product and under what conditions.
A Little Negotiation
If you've requested that a list owner or manager suppress a certain data variable, it's never a bad idea to have your service bureau put on an extra screen for that variable when it runs the merge-purge on your file, says Wheaton. If you work out a deal with the bureau where you only pay on a per-hit basis, then you can take the hits back to the list owner or manager and request a rebate on the names it was supposed to have suppressed.
Think this is excessive? Think again, says Wheaton. "We have a client who for years on all list orders specified 'no company names.' When we started working with this company, we added to the merge/purge instructions for the service bureau to suppress company names... We found out that the vast majority of list owners had never deleted company names "or weren't doing it correctly. We had hundreds of thousands of company names [in the rental files]." What's more, Wheaton found that some of the lists performed well with the company names left in.
The above example is the reason why Goldberg always eyeballs a sample dump from every list his clients mail.
A final tip: Tremaine adds that it doesn't hurt to explain to list owners and managers your need for a net-name rate other than the standard 85-percent. If you know that your dupe rate will be a lot higher than I5 percent, it's wise to negotiate a new rate so you don't pay for names you already have as customers and avoid inflated run charges on large files.
