How to Improve Upon Generic Cooperative Database Response Rates

A big issue with catalogs is the need to expand their profitable prospecting universes by improving upon the pure vanilla models served up by the cooperative databases.

Here are some proven tactics from Jim Coogan to improve the performance of cooperative database models.

1. Find preselects that improve performance. Preselects are additional variables that can either find the best names in a model or eliminate poor responding names.  Common preselects include geographic or zip models or demographic overlays.
2. Use multi buyers from multiple coops and vertical lists. Measure multi buyers and singles separately.  Suppress singles if they underperform.  Negotiate with the secondary coops to add names for building larger pools of multi buyers.  Multi buyers simply work better than singles.
3. Score the prospecting names using cooperative database optimization modeling and drop unprofitable singles and multis.
4. Build affinity tables. Affinity tables are the lists of catalogs who should be contributing to a coop database models as well as the lists of individual catalogs that should not be included in a coop databases models.  The catalog owners knows his own competition better that the analyst at the coop and the cataloger should be deeply involved in defining the affinity tables of which catalogs should be contributing to models and, sometimes more importantly, who should not be contributing!  List the catalogs whose buyers you don’t want to prospect to.  Build your own list of catalogs that you should be prospecting to and ask the coops to give you buyers from those catalogs.  Giving the coops an affinity table of catalogs most likely to buy from your catalog really helps the coops build models.  Share the universe of vertical lists that work with the databases and also share the names of the vertical lists that fail.
5. Break down large models into smaller segments so you can measure where performance declines below breakeven.
6. Model just your “fall season buyers” or your “holiday buyers” for a fall or holiday season mailing. This is especially useful if you have different buyers for different seasons.
7. Model just your very best buyers. Take your 3x buyers (if you have over 5,000 buyers) and model those best buyers.
8. Compare your best buyers to your entire buyer file and see if the differences between your best buyers and your total buyer file lead you to a good preselect.
9. Consider dropping the low ticket (<$25 buyers or <$35 buyers) from your modeling especially if the low ticket buyers aren’t proving profitable.
10. Model just your recent new-to-file buyers. If your market is pretty mature, this may point to where you are able to model effectively for new buyers.
11. Test relentlessly. Do stronger offers raise prospecting list performance? Does seasonality play a role in prospecting performance?  Could lower page counts and catalog costs improve breakeven without cutting into response rates?  Know the outside factors that affect the response rates of your prospecting circulation.
12. Segment and test “previous” versus “new” names in your coop models. Usually “previously mailed names” respond better because these are mail order buyers who a) consistently rise to the top of your models and b) have received the catalog in the immediate past so they have read about and know your brand.
13. When you test an nth selection of a model, take the best names from the model rather than an nth selection across the entire model. This maximizes your sales.
14. Work directly with coops and meet with them constantly. Visit Denver and you’ll get face time with your day to day contact, management and the analytical and statistical people behind the curtain.  The coops have really smart people working for them and the more they understand your business, the more solutions they can provide to building a bigger profitable prospecting universe.
15. Get the coop’s free demographic reports showing how your own customer’s index according to demographic and transactional characteristics. You’ll get some startling insights into your own customers and into ways to improve your prospecting!
16. Ask if the coops can provide a segmentation of modeled names with and without a “cookie.” Hint:  names with “cookies” are more savvy buyers.
17. Use the hidden gems of coop database house file optimization to reactivate older buyers, ship-to’s, catalog requests and giftees. The coops can tell you precisely which of these older house file names can be mailed profitably.  Also use the coops to optimize vertical lists and magazine lists to find the profitable circulation.
18. Do you have disjoint groups in your house file? Disjoint groups are a statistician’s term for different kinds of buyers.  Examples could be horse owners versus people who love western wear, or Affiliate buyers versus catalog buyers or web buyers who came to you and didn’t receive a catalog versus buyers who came from prospecting or retail buyers versus catalog buyers.  If you have disjoint groups, model those groups individually and you’ll get better results.
19. Are your “Pure Web Buyers” who came to you through the web and never received a catalog different than your “Catalog Buyers?” You may need to model the “Pure Web Buyers” and the “Catalog Buyers” separately.
20. Segment the BTB buyers and the BTC buyers. Can the coops provide BTB models?  Should you be sending more prospecting names to BTB addresses to expand the reach of your catalog to other prospects within a business?
21. Consider flagging prospecting names and suppressing names that you have mail a prospecting catalog to 10, 15 or 20 times. While these names may float to the top of the models, the names that you have mailed more than 10 or 20 times in the past are unlikely to make a purchase.
22. The coops provide digital ad campaigns. They are getting really, really good.  Why?  Because the coops understand the metrics and the profitability needs of catalogers.  Because the coops understand the basics of direct response.  But the real key is that the coops can provide pools of proven prospects because they are built using databases of behavioral and transactional data.  The coops can serve up households that are currently actively purchasing within your merchandise category!  And the coops can cookie your house files and prospecting mail files and provide ads to those households that have just received your catalog. Digital ad campaigns have proven profitable ROI.
23. Coops can always provide incremental pools of retail names if you have retail stores. And if you have certain geographic regions that index higher, prospect deeper into those cities or regions even if you don’t have a retails store.
24. Look at your match backs and understand how many sales are unallocated and how much your raw prospecting results should be increased by allocating a portion of the unallocated sales to your prospecting. Your “raw” prospecting results are always somewhat understated. The question is how much should you add to the prospecting “raw” results to read your true results.

PS.  One old coop database owner took me aside years ago and reminded me that if I didn’t know anything else to remember that “Pre Selects” were the key to improving performance.

PPS.  Visiting the coops regularly to review results is the most efficient way to find ways to improve your results!  Face time works.

Jim Coogan

Catalog Marketing Economics

505 986 9902 office

505 699 2948 cell

jcoogan@earthlink.net

1328 Bishops Lodge Road

Santa Fe, New Mexico 87506

Building Mail Order Circulation Profitably