How to Get the Most From Cooperative Databases

Co-op databases provide a valuable source of quality names for many product offers. For most small- and medium-sized catalog companies, 50 percent or more of prospecting circulation goes to names selected from such databases. Prospect names selected from co-ops are a good value for the money.

Why? First, the names are selected according to a model, which means they resemble your own customer database. Second, the names rent for less than most outside response lists. They’re “net” of your housefile, which translates to a lower cost per net name. Co-ops may appear to be about half the cost of outside lists, but in reality, names from a co-op are about one-third the cost, again because they’re net of your housefile.

Some co-ops can base models on variables created from SKU-level data — a real advantage that’s reflected in the revenue realized per catalog mailed. Each co-op starts its model from a different database structure, resulting in unique variables to study. And each co-op creates its models using different methodologies. So while there’s a significant overlap of names through co-ops, the modeling techniques they use can select different individuals for mailing. One co-op may identify good prospect names that another co-op missed. The objective in working with multiple co-ops is to find those hidden prospects.

To get the most out of co-ops, work with them as a partner as well as an extension of your marketing team. Share your goals and results with them. Co-ops supply prospect names. But the quality of the names obtained can be enhanced by working more closely with them, and their resources can be applied creatively to improve your housefile and prospect-name performance from vertical lists.

Seven Proven Strategies

• Outside list optimization. All mailing lists contain some households whose overall buying patterns aren’t typical of the other households on that list. For example, a high-end gift list may contain prospects who bought for a special occasion, but generally buy from lower-end catalogs. Optimization is the process of matching vertical list names against a co-op database and scoring them with co-op targeting tools. This identifies households that don’t otherwise fit the desired buying patterns. These less-productive names then can be suppressed from the mailing, and the response rate from a particular list will improve as a result.

Similarly, this technique can be used to optimize the performance of marginal lists by suppressing the weakest segments — those that perform just below acceptable response. Using this technique typically suppresses 10 percent to 15 percent of a list, thus providing enough of a lift to make a marginal list profitable or to help weaker segments achieve breakeven. Usually the lift in response more than justifies the incremental cost of optimization.

• Suppression of rental singles. The most common use of list suppression is optimizing the rental singles (one-time buyers) who come out of a merge. The worst-scoring segments of the rental singles frequently perform at less than half the response rate of the average rental name. In many cases, mailers replace the suppressed names with better prospect names. This process allows you to improve the overall performance of the outside names mailed and to fine tune the number of names on the mail tape sent to the printer.

• New-to-file buyers. Mail new buyers faster by selecting recent fill additions from the co-op database. Set up a weekly or monthly program (depending on volume/quantity) to download these hotline names for mailing.

• Housefile modeling. You can use housefile modeling for specific applications to improve your revenue per catalog (RPC) mailed, but not as a substitute for recency, frequency and monetary value (RFM) analysis.

One effective use: Model all the previous buyer names you don’t plan to mail, including older names for reactivation and non-converting inquiries. Using co-op models to reactivate your housefile will help identify previous customers who are most likely to buy again. The same is true of using a model to identify which inquirers are most likely to buy. In both cases, you can increase RPC by eliminating circulation to non-performing names before mailing them. After your final mailing to non-converting inquiries (generally mail three times), have these prospect names modeled, and mail them one more time. You may see an increase in your RPC.

• Add a mailing. If you can squeeze in one more mailing to your housefile and to the best performing names from a co-op database during your busy season, do it. Let’s assume you mail to your housefile 12 times a year (every month). Next, assume the winter holiday season is your best time of year. During September, October and November, mail three weeks apart vs. four weeks to mail one more time to your housefile. This additional drop should perform at 70 percent of a regular housefile mailing to like RFM segments of your customer list.

• Leveraging SKU-level data for modeling. SKU-level data not only offers greater precision in categorizing buyers, but it also can be used to build custom models designed to clone item-, category- or season-specific buyers in a way that marries the analytic tools to your merchandising strategies. It’s sometimes more difficult for a catalog to provide SKU-level data, but the superior results justify the extra cost and effort.

• Enhanced data overlays. External data can enhance results with co-op databases. For example, one co-op offers attitudinal segmentation through its partnership with Yankelovich and its MindBase segmentation system. This unique partnership allows its members to link behavioral targeting with the underlying attitudes and motivations driving purchase behavior. Results to date show this extra dimension offers significant potential in opening new co-op universes for catalogers, as shown in the chart Yankelovich Model I-Behavior Test. In the test, the use of the Yankelovich model for this food mailer significantly increased the response rate and RPC. This enhanced data overlap model has proven to be more effective with niche offers. (Results will vary, which is why testing is critical.)
Conclusion

Some catalogers have a problem joining co-ops (or more than one co-op) because they feel a need to protect their customers’ names. Since co-ops don’t release names that are unique to your company, this shouldn’t be a barrier for joining. Your names are being mailed other offers regardless of your participation in the co-ops. You may as well take advantage of co-ops as another source for names.

Some co-ops supply customer profiles and/or demographic information on your file, which can be extremely beneficial. They can provide data that are more valuable than expensive research studies.

Stephen R. Lett is president of Lett Direct Inc., a catalog consulting firm specializing in circulation planning, forecasting and analysis. He can be reached at (302) 541-0608 or by e-mail at steve@lettdirect.com.