Perfecting an Imperfect Science

With the Internet fast becoming the vehicle of choice for catalog shoppers to place orders, catalogers face the challenge of identifying what inspired those buys to begin with. The more marketing channels available, the more ways companies can reach existing clients and prospects. But unless a buyer purchases the old-fashioned way — via phone, fax or mail — or types the key code of a catalog into the Web order form, it’s tough to determine whether that sale was driven by the catalog, an e-newsletter, a Google search or an intentional visit to your site.

Earlier this year, Catalog Success polled multichannel merchants to show the differences in how they handle matchbacks. Only a little more than half of those surveyed have matchback programs. And among those that do, just 41 percent said their programs are accurate and reliable.

Duluth, Ga.-based National Allergy Supply conducts matchbacks internally. John Fry, vice president of marketing and sales, feels that some of the conclusions one may draw from the data can be subjective.

“It’s accurate as far as confirming who received a catalog and whether they placed an order,” he says. “Many catalogers say, ‘If you received a catalog and you placed an order online and you’ve placed orders online before, for the sake of response rates we’re still going to believe that the catalog influenced that order.’”

Could one also assume that 25 percent of those customers threw the catalog away and weren’t influenced at all? “All you know,” Fry says, “is that they got it and that they placed an order.”

Matchbacks help Fry and his team on a more concrete level with square-inch analysis. “Having that information about customers’ orders still allows us to determine what’s selling in a catalog, and what’s not,” he says.

Monitor Response
Lorraine Calder, president/CEO of Litchfield, Conn.-based White Flower Farm, says that for the past two years her company has been using an outside service to produce weekly matchbacks. In Calder’s business, the catalog is the driver, and she matches back to the most recent mailing.

“The information is very accurate, assuming that the customers are the people whose names and addresses are on the catalog,” she explains. “If they move or get married and I don’t know that, that’s different. But for the most part, we’re tracking them pretty closely.” She also notes that this information helps her monitor response by product segment.

Until recently, Richmond, Va.-based children’s apparel cataloger CWDkids had been matching back twice annually through a co-op. But at press time, Tracy Schneider, vice president of marketing and operations, said she was awaiting the results from a first matchback conducted by Web development provider iCentric. “We took an entire season and allocated based on order curves,” she explains. “We found a buyer of ours may get multiple catalogs during the season, and we just had been matching them back to that one catalog twice a year.”

Now CWDkids examines how many times it mails to specific customers. And with the matchback data, it’s determining whether it can cut back on the number of times it contacts customers to better control both mailing and paper costs.

“We’d also like to incorporate our e-mail contacts into the matchback process so we can tell if we’re also contacting these customers via e-mail,” Schneider says. “Are they responding to the e-mail, or are they responding to the print catalog? The more information we can get from the matchback, the smarter we’ll be able to mail in terms of both e-mail and catalogs.”

Dental products cataloger Practicon approaches matchbacks by analyzing purchases made from customers already in its database. “From there,” says Marketing Analyst Stan Moore, “we look at how they’ve ordered before and what channels they’ve ordered through.”

Things get a little hairy for the Greenville, N.C.-based Practicon when new customers place orders. For one, he doesn’t know how they come to the company. “They often come to us through lists we’ve purchased or rented from our broker” Moore says. In this case, the matchback is handled by the company’s outside data warehouse. “There, we’re just matching back on the address that’s come into our system, as well as having showed up on the rental list.”

Not Gospel
Debra Ellis, president of Barnards-ville, N.C.-based Wilson & Ellis Consulting, views matchback data as “interesting information,” but not gospel. “Matchbacks are a good guess,” she says, adding that if catalogers place too much weight on them, they could make marketing decisions that are detrimental to their businesses.

“I have seen people go in and substantially cut back on their catalog and direct mail pieces thinking that everything’s being driven by e-mail, because that’s what the matchbacks are showing,” she points out. “Then six months later, their sales are down. When they beef up catalog and direct mailings, sales begin to pick up again.”

While National Allergy Supply’s Fry notes that the information gleaned from matchbacks is useful, he’d like to validate, for certain, whether his team is making the correct assumption on those who are influenced by the catalog.

“We’d like to learn whether we should continue mailing catalogs to them,” he says, “and about how to use the customer purchase data to seamlessly and easily provide offers through targeted e-mail programs to these cross-channel customers.”

Moore points out that Practicon’s growing online business raises the issue of how to pinpoint what drove customers to the Web site to begin with. “With an Internet sale,” he says, “we’re struggling to know if it’s come in via search, a catalog or just a visit to the site.”

At White Flower Farm, the matchback challenge lies in seasonal overlap. Being a gardening business, seasons are very distinct; however, customers don’t always see it that way. “The minute my fall book goes in the mail,” Calder says, “it’s getting credit for the order, even though it may be a spring product. That’s where I have some muddy water, and I don’t know that any matchback tool can fix it.”

When analyzing matchbacks, Ellis says she prefers to track one customer’s purchase history to identify how many times that customer was contacted before actually making a purchase. Then she explores what medium the customer used when placing an order.

Matchbacks, however, should remain just one element in the entire multichannel marketing strategy. “Look at the whole picture in your marketing campaign to understand where it’s coming from,” Ellis says. “Matchbacks will show one contact point that may have contributed to a group of customers or made a specific sale to an individual customer.”

Carolyn Heinze is a Vancouver, B.C.-based freelance writer/editor. You can reach her at carolynheinze.blogspot.com.