Link This | Email this | Blog This | Comments (0)

Clean Data Make All the Difference in a Mail Campaign
August 15, 2007

Last week on Marking Marketing Work, Heidi Tolliver-Nigro wrote about the benefits of highly targeted direct mail. And, of course, one of the most important prerequisites of targeted marketing—whether it is print or electronic—is good data. That is, a clean, well-ordered database.

A clean database serves many purposes, the two most important being to cut costs and increase effectiveness. For example:

• Getting rid of duplicates and bad/out-of-date addresses can cut mailing costs by eliminating “useless” printing and postage.
• Spelling a potential customer’s name correctly or getting the name of their company correct, especially when  using targeted marketing, can improve the effectiveness of a mailing by not calling attention to the fact that the sender is using an old database.

In terms of the former point, we just received data from The Industry Measure’s Summer 2007 survey of publishers and the number one action both magazine and catalog publishers have taken vis-à-vis the recent increase in postal rates has been to clean their mailing list—that is, purge duplicates and bad addresses.
As to the latter point, I have often received a targeted direct mail piece on which my name was misspelled, or the company name was one I have not worked for in years, calling attention to the fact that the sender was using an old list—shattering the illusion of “personalization.” In fact, seeing bad data used in a way that is meant to impress us ironically imparts more of a sense of “impersonalization,” obviously not what the sender intended.

And it’s not just print. I occasionally receive e-mail pitches from online retailers (such as Amazon) that are reminiscent of what I call “When Databases Attack.” For example, a headline may read, “You recently bought Citizen Kane on DVD. We thought you might like to know that other customers who have bought Citizen Kane also bought Plan 9 From Outer Space.” Naturally, Orson Welles’ Citizen Kane—recognized by critics as one of the best movies ever made—has nothing to do with Ed Wood’s Plan 9 From Outer Space—recognized by critics as one of the worst movies ever made. This is an extreme example, but not far afield of some actual e-mails I have received.

How does this happen? One wonders, but likely one was bought as a gift and one was not. Perhaps someone is studying cinema and wants examples of extremes of filmmaking. Maybe one particular customer has really eclectic tastes. Who knows? There could be a million reasons why customers would buy such disparate items, but the point is that a retailer’s content management system can’t make such nuanced distinctions. Thus, the retailer likely doesn’t sell anything, but instead makes his e-mail pitch the butt of a joke that may end up on that target customer’s own blog. Trust me…

So while databases are excellent tools for marketing, especially personalized marketing, we need to be cognizant that pitfalls abound. Obviously, it’s impossible to devote time and resources to verifying every piece of data, but there are some steps you can take to ensure that your marketing efforts are not at best ineffective or at worst laughed at:

• Do a broad, static (that is, non-variable) initial mass market mailing to your list, and if any return, use them to update the database.
• Work with a mailhouse that uses CASS (Coding Accuracy Support System) Certified address matching software, a mailing list technology that standardizes addresses and matches them to the USPS’s own database, thus validating delivery address information and identifies inaccurate, incomplete, or incorrect addresses.
• Try to spot and correct simple typos—“Rihcard” instead of “Richard,” for example. (Typos like this can also make it easy for recipients to know where certain senders got a list.)
• Pull a sample of the records in your database and design your marketing piece around those names. Do some names use abbreviations that would look weird if used on a variably printed piece? Are there other considerations withy regard to the data that would make a certain design approach work less well than another approach?
• Try to inspect a representative sample of variable printed direct mail pieces before they go into the mail stream—preferably before they get printed (front ends often allow you to see all the variable pieces onscreen in PDF format).
• Likewise, inspect a representative sample of electronic pitches, as well. Bad e-mail addresses aren’t as big a problem as bad snail mail addresses (bouncebacks at least don’t waste postage) but other content-related issues may blunt the impact of the marketing pitch.

A few common sense actions can make the difference between a highly effective or a highly ineffective mail campaign.

Posted by Richard Romano on August 15, 2007 | Comments (0)