Steven Hulet and Dr. Sean Warnick, Computer Science

Because of its relationship with Brigham Young University, the BYU Bookstore is in a position to know more about its customers on an individual basis than possible for the average firm operating in the free market. Such in-depth knowledge of a customer base allows a firm to better serve its customers and can result in greater satisfaction for the customers and greater revenue for the firm.

Every student at BYU is already uniquely identified by their student identification number. The Bookstore already has the capability of capturing this number, along with other transaction data, when the student ID card is swiped as part of a transaction. Our goal became to create a program encouraging students to allow their ID cards to be swiped as part of each Bookstore transaction, giving the Bookstore access to transaction-level data specific to each customer. To encourage participation, we planned to offer an incentive or reward for each swipe.

The large amounts of data involved make this an ideal application for computers; which data are most relevant and how to extract information from the data are two questions often dealt with in computer science. Issues such as data structures, algorithmic space and time complexity, and data representation will all need to be dealt with.

Programs of this nature have become a trend in recent years, known as customer loyalty programs, reward programs, or frequent shopper programs. Ours was to be different in several respects. Our goal was to collect more demographic data on our participants than most commercial implementations, thus making it easier to extract meaningful information from the data and identify market segments. We planned to do this by looking at individual people and individual transactions, something else rarely seen in practice due to the computational complexities involved. Perhaps most notably, the Bookstore was willing to allow us to conduct controlled experiments, varying prices or introducing targeted promotions, and study the responses. This kind of real-life laboratory is nigh impossible to achieve with any other firm, but would allow us to verify the accuracy of our results.

We presented our ideas to the Bookstore and they agreed to help set up such a program. They agreed to allow us to analyze the collected data and were excited by the information and advice we could potentially provide them. In designing the program we consulted both the published literature and the Bookstore management staff.

Based on existing research we classified customer loyalty programs into various types. As reported in Ran Kivetz’s The Effects of Effort and Intrinsic Motivation on Risky Choice, he type of participation incentive could be small-sure or large-uncertain. A small-sure reward ensures a small guaranteed payoff at guaranteed intervals, such as a small fixed percentage discount on all purchases. A large-uncertain reward is one which is not granted to everyone, but as a corollary, can be significantly larger when it is awarded. A cash lottery is an example of this. Customers tend to prefer small-sure rewards when they perceive an effort required to participate and largeuncertain rewards when there is no perceived participation effort.

The Optimal Choice of Promotional Vehicles: Front-Loaded or Rear-Loaded Incentives? by Zhang, et. al, addresses the differences between offering front-loaded incentives, where the incentive is realized at the time of purchase, and rear-loaded incentives, where benefits are realized at the next purchase or later. These are the differences between peel-off coupons and inpack coupons, or between an up-front percentage discount and accumulating points toward a percentage discount. Firms achieve maximum profit by offering front-loaded incentives if their customers tend to each buy a few products on a regular basis; firms serving customers who like to buy different products with each visit are best served with rear-load incentives.

Another important consideration is that different rewards with the same perceived value to the customer may incur greatly different costs to the firm. A $5.00 cash reward and a $5.00 gift certificate both have a $5.00 value to the customer, but the cash could be spent by the customer anywhere. A gift certificate, on the other hand, assures that the monies return to the firm, and the firm loses only what it paid for the items purchased. Another option is for a firm to offer goods or services for which it normally charges, yet which actually cost the firm itself little. We presented these considerations to the Bookstore management team, and heard their input as to the nature of their clientele. We together designed a program which should appeal to everyone. Our program would combine a small discount for students with each participating purchase, and a free Bookstore service with every fifth participating purchase. Additionally, lotteries for larger prizes would be occasionally held. Students would be entered each time they allowed their ID card to be swiped in conjunction with a purchase.

As we continued the program design more people became involved. Communications and marketing experts were brought into the project to address questions such as the program name and logo. This led to additional discussion and debate, and resulted in the target launch date for the program being delayed. The most recent timeline we received calls for a three year period of study by the communications and marketing people to fully understand the customer response to the program and the impact it will have on the BYU Bookstore brand image.

Nevertheless, although the customer loyalty program remains under development, the data analysis project continues. Instead of operating on the transactions data from the entire BYU Bookstore clientele, we are continuing by focusing on a subset of customers—those who shop via www.byubookstore.com. The Bookstore web site already captures transaction level data for thousands of customers, with records going back around three years. By analyzing this data now, and more in the future, we will learn more about the customers of the BYU Bookstore and design methods of better serving them.