Loyalty Programs and Data Mining

By Meta S. Brown

A loyalty program is an agreement between a business and its customers. Customers agree to allow the business to track purchases (and possibly other actions as well), and in return, the business offers rewards. Typical rewards include lower prices or a free product or service.

You may be involved in several loyalty programs as a customer right now. Airline frequent flyer programs are loyalty programs. So are wholesale club memberships, preferred shopper cards, and even coffee shop punch cards.

Every loyalty program depends on the cooperation of the customer. The customer must first opt to participate in the program, and then follow up on every transaction. In the simplest cases, the customer might carry a simple paper punch card, which the business would mark each time a purchase is made.

When the customer has made some required number of purchases (perhaps 10 or 12), the card can be handed in to the business in exchange for a free item. But punch cards don’t provide the kind of information you need for data mining!

More sophisticated loyalty programs provide the customer with a card resembling a credit card that can be electronically scanned. When the customer forgets his card, he may have the alternative of giving a membership number, phone number, or name to locate the proper account.

These programs are important for data miners, because they allow the business to track the customer’s purchasing behavior in detail. Some businesses use this type of tracking specifically to get data for analysis. But others have a different motive: Paper cards and coupons can be easy to counterfeit. So some businesses choose these computerized tracking methods primarily to protect themselves from losses due to fraud.

Smartphone applications allow customers to identify themselves and receive coupons and other offers without the need to carry membership cards. These applications are popular with shoppers who find them convenient.

For the retailer, smartphone apps can provide a depth of information that isn’t available through any traditional loyalty program, including real-time browsing and geographic data. Smartphone-based loyalty programs also facilitate sharing by users. It’s easier to share a great promotion on social media with your phone than to pass around paper coupons, for example.

Online retailers have special advantages. Their customers usually set up accounts when making a first purchase, and provide information that can be used for tracking, such as an email address. These retailers don’t necessarily promise rewards for setting up an account, yet they can collect data on transactions and on all sorts of online behaviors.

They can also track visitors using weblogs and cookies, tracking information associated with web browsing. They know when a customer visits the website and what products the customer views, even if no purchase occurs. Users may voluntarily add additional information to their own accounts, such as product reviews, wish lists, and user profiles. Every bit of this information has value for data mining.

Here are some of the data elements that may be available to you as a data miner in the retail sector:

  • Customer location

  • Products purchased

  • Combinations of products purchased together

  • Prices paid

  • List or everyday prices (often different from the price the customer paid)

  • Coupon or other discount offer used

  • Time that a purchase took place

  • Detailed product descriptions

  • Pages/products viewed

  • Time on site

  • Timing of site visits

  • Product reviews and information sharing

  • Referrals (for example, did the customer come directly to the site, or via a link from another site, or a link in an email?)

  • Offers or ads that the customer viewed

  • Social network details, such as people the customer knows

This information is a treasure trove for marketing! As a data miner, you could hardly dream of a more valuable data source to help a business understand what and how to sell to individuals. But be wary. Boundaries exist on the appropriate use of this personal data. Some are defined by the law, and others by public sensibilities and the preferences of the individual customer.