Listening to your customer (data)

By Published on .

How are recommendations working for you? Amazon attributes a whopping 35% of its product sales to recommendations. While that may be quite a number to fathom, it’s not as difficult or expensive as you might think. Data-driven marketing applications are practical, affordable and something you must consider as part of your marketing mix. Let’s take a look at creating personalized recommendations that maximize the value of current customers through a thorough analysis of your existing database.

What do your customer data say? The first step is to organize and review your current customer list. Most marketers segment customers into deciles (tenths) in terms of their RFM (recency, frequency and monetary value). Look for patterns such as percentage of revenue, average order size and behavior comparisons between the top 20% and the balance of your database.

The second step is to perform additional analysis to predict customer tendencies in terms of clusters and other data-modeled groupings. This will help you determine the type of message or offer to send each group based on past behavior.

Now you are ready to leverage your data and analysis in order to move your best customers to higher purchase levels. This might take the form of making a specific set of cross-sell or up-sell recommendations to individual sets or clusters of customers. For example, if you are a kitchenware manufacturer and your analysis shows that customers who purchased high-end blenders also frequently purchase espresso/cappuccino machines, make a special offer to those blender buyers picturing the latest and greatest espresso/cappuccino machine.

If your “house” or current customer files are not in the condition necessary to take advantage of some of these types of analyses, make it a priority to clean up your data so you can maximize your ability to harvest your own greatest asset.

Specifically for you
When you have a solid set of customer data, you can take advantage of a technology such as such as variable data printing (VDP), which allows you to reach your customer with a uniquely personalized, printed direct or electronic mail piece. VDP mailers can be created using a set of rules to swap images (from an image bank) based on information pertinent to individual data records. This technique can really help increase sales to existing customers.

Travel companies have successfully used VDP, for example, to suggest new vacation locations to professional customers in a particular income bracket. So if a professional living in New York last visited Hawaii on vacation, and the travel company database analysis showed that similar customers next vacationed in the Caribbean, a postcard or other mail piece might go out showing pictures of locations in the Caribbean.

If you combine this with a follow up e-mail campaign using the same look and feel, you have a coordinated, targeted marketing effort that should deliver an excellent ROI. An interesting note: VDP appears to work best with larger-ticket items, as the individual cost on a per-piece basis is sometimes three or more times higher than the cost of a static printed piece.

May we make a recommendation?
What about recommendation systems? The idea is that customers are not all the same so why would you treat them that way? Amazon was the first to really use the “Customers who purchased this also were interested in the following items” technique. But you can use the same approach for your own Web site.

Personalized recommendation systems can cost more than $100,000 for companies with a large number of product offerings. But for companies with a smaller amount of products, costs can be less than $50,000 for set-up, and monthly maintenance charges can be tied to performance.

An added benefit is that creating a truly personal shopping experience for your customers engenders increased customer loyalty. We return to Amazon when looking for items for ourselves and gifts for friends and relatives. The friendly service seems to know us—it certainly recognizes us—and it appears to have been waiting patiently for our return. Why not create a “my store,” “wish list,” e-mail alert or an intelligent search function on your site to make your customers’ shopping experience truly unique? Recommendations can be made such as “No. 1 in its category,” or you could promote sale or overstock items. Today’s shoppers want to feel that they are in control of their shopping experience and affording them choices to make (but not too many) is a key to success.

So whether it is in print, on your Web site or both, data driven applications offer your precious customers new products, destinations and experiences—and provide you with stronger income opportunities.

Mark Kolier is president of direct marketing agency, CGSM (

Most Popular
In this article: