Several of the cooperative databases collect product-level data. Marketers are finding product-level data very useful in data mining for customers who buy specific products through direct marketing.
But using the co-ops’ product-level data isn’t the only way to get at customers who buy within their product niche. Here are three alternatives:
- Use all the buyers of certain types of products and model only those buyers. This has the advantage of allowing you to build based on very specific selections of products rather than broad product selections offered by the co-op database. Building models based on a cataloger’s specific responder file is a proven model building strategy.
- Build a table of catalogs whose product lines resemble your overall merchandise mix or are like a specific subset of your merchandise mix that you want to target. Telling the co-op database to model based on purchase activity from specific catalogs works. Building tables of catalogs to not model and excluding certain catalogs’ transaction data is also a powerful model-building tool. If you can identify certain catalogs that closely mirror a subset of your merchandise, build a table of those catalogs. If you can identify certain catalogs whose buyers you don’t want to prospect to, build a table of those catalogs and exclude their transactions.
- Model the product buyers from your buyer file that you want to prospect to. Compare the demographics of those buyers with the demographics of your overall buyer file. The differences between the overall buyer file and a subset of buyers may show you selects you can use to zero in on your best prospects or to exclude prospects outside the “sweet spot” for specific product categories.
Product-level models can be a powerful tool. But they work only if the co-op databases have product selects that fit the merchandise category you need. If the product categories are too broad or not well targeted to your merchandise, product-level modeling may not be your best solution.
Make sure you look at all the modeling solutions that let you narrowcast your modeling toward the merchandise niches you want to grow. And be certain to test product-level modeling against these other techniques to get the biggest, most profitable prospecting list universes.
Jim Coogan is president of Santa Fe, NM-based Catalog Marketing Economics.