Game, Set, Match-back

Most catalogers have long used source codes to determine the success of the lists they’ve rented. Codes indicating the source of the recipient are printed on the catalog cover and the order form; call center reps ask for the source code as part of the order-taking process.

But orders placed on the Web are trickier to track. Even if customers were driven to the Website by a catalog mailing, they may not have the book — and their source code information — at hand when placing the order. Sandy Matika, senior vice president of Hackensack, NJ-based list firm Mokrynski & Associates, says that only 20% of the clients she surveyed are matching back their Web orders to a source.

Failing to match back Web orders can cause you to dismiss effective rental lists as failures. “I have lists where sales driven to the Internet add a 15% lift. I also rent lists that generate an additional 80% of demand on the Internet,” says Linda Spellman, director of direct marketing for Corte Madera, CA-based home decor cataloger/retailer Restoration Hardware. “Without matching back the Internet sales to those lists…you’d miss that.”

“By matching back Web sales to circulation you can see the value in lists,” Matika says. “You could skew your making circulation and budgeting decisions if you’re not doing it.”

Capturing the codes

Key to matching back Web orders is capturing source codes. One way to improve your capture rates is to add a pop-up screen during the checkout process to remind customers to enter a catalog source code. Explain that you want to accurately gauge where your customers are coming from. “If you give customers a reason that you need their information,” Matika says, “you’ll be able to bump up your capture rates.”

As they check out from the Day-Timers Website, for instance, buyers come across a box for the source code with copy that reads “Be sure to enter the source code to get the latest catalog offers.” If the customer does not enter a source code, says Tom Tweedie, director of catalog and Web marketing for the East Texas, PA-based marketer of productivity tools, a default code goes into the system attributing the Internet as the source.

Incentives can boost online source code capture rates too. You might offer customers a $5 discount toward their next purchase in exchange for source code information.

Langhorne, PA-based gifts and collectibles cataloger Lenox Collection gets an assist from Web surveyer BizRate. Lenox contracts with BizRate to survey its online customers regarding the shopping experience, explains Maria Youth, vice president of catalog and Internet for Lenox. But the BizRate survey also asks buyers what drove them — a catalog, an online ad, a magazine ad — to the Website. Lenox markets not only by catalogs and the Web but also via solo fliers and space ads in magazines such as Parade.

Of course, the more channels, promotional campaigns, and prospecting sources you use, the more complex matching back is. Unless all of your campaign data and customer database history are fully integrated in one file, “doing match-backs yourself is virtually impossible,” says catalog consultant Coy Clement, founder of East Greenwich, RI-based ClementDirect.

Even the simplest method — taking all of your Web customers and running their names against the names of your most recent catalog mailings — usually requires assistance from a service provider. According to Matika, most catalogers use the company that performs their premailing merge/purge to conduct this sort of match-back. The cost is $500 to $1,500 per project.

Complicating matters

Just because a buyer received a print catalog doesn’t necessarily mean that the catalog — and the list from which the name was rented, if any — should receive credit for the sale. Did the customer also receive a promotional e-mail, for instance? Or perhaps he came to the site from an online affiliate’s site. Ideally you need to view data on all media that could have impressed upon the customer.

Several software programs have evolved in recent years to help catalogers do just that. Co-operative database services provider Abacus launched its ChannelView program in February 2002; database services provider Experian debuted Channel Match in May 2002. Both systems use the cataloger’s promotional files and response data from mailings and Web data. Channel View is Web-based, Channel Match is not, but soon will be. Costs range from $10,000 to more than $30,000 annually, depending on how many catalogs you mail annually.

But software packages aren’t infallible. Lenox’s Youth says that they tend to assume that the catalogs are the drivers to the Website. While the system’s algorithms allow mailers to segment specific channels, the software’s logic uses catalogs as its default mechanism. And if you mail more than one title or market by more than just a catalog, those results may be misleading.

The Rules of Matching Back

Remember the good old days of order trackability? Those days are gone — Internet orders are increasing and seldom have codes, customers now receive e-mails that overlap with catalog mailings, and more catalogers are opening stores, adding to the confusion. That’s where order allocation comes in. Coy Clement, founder of East Greenwich, RI-based catalog consultancy ClementDirect, explains his match-back philosophy:

  1. Matching back is a critical business issue, not a technical detail. Not performing match-backs or doing it incorrectly can lead to bad business decisions. All of a company’s senior management should be actively involved in match-back decisions.

  2. Keep an eye on total dollars and orders being reallocated to different sources. Are the percentages increasing? Are they so material that small changes in method could result in serious changes in business decisions? For instance, would reallocated dollars cause segments or even entire mailings to drop below breakeven?

  3. Understand that all match-back decisions are based on assumptions — for instance, how many days is the “life” of an e-mail vs. a catalog. How these assumptions are handled can change mailing decisions.

  4. Some offers, such as discounts, are more trackable than others. Delivering special offers mostly via e-mail rather than the catalog creates a higher percentage of online orders and makes print look less profitable. Is that really good for your business?

  5. Run periodic, highly trackable promotions that can be redeemed in all channels. This allows you to measure how customers are actually using channels and provides a “check and balance” on allocations.

  6. Monitor whether e-mails are working the way you think. Are customers who click through to the site buying the items or categories featured in the e-mail? If not, should the e-mail receive credit for all the orders? You want to know if e-mail works better because customers also receive a catalog. If all the orders are allocated to the e-mail, some catalog segments may no longer be deemed profitable.

  7. Look at alternatives to your match-back assumptions. For instance, if all orders from customers receiving an e-mail are allocated to the e-mail for seven days after a mailing, what would the difference be if the window were reduced to three days?

  8. Order allocation is art as well as science. A highly seasonal catalog may want to allocate orders differently for customers who receive multiple issues in a season. One method is to spread the reallocation to all catalogs in the season rather than to the mailing closest to the order; this avoids concentrating all the circulation in a single issue and minimizes risk.

  9. Periodically validate that match-back is working the way you thought it would. Schedule reviews at least once a year to monitor the process: Has the percentage of “unknown” sources changed? Have you increased or decreased the frequency of catalog mailings or e-mail messages?

  10. Pick the frequency of match-back that works for your business. For some businesses, such as those with retail stores, daily or weekly match-backs are important. In others, performing match-backs once a month or even once a season may be adequate.