Doctor of Philosophy
Evaluating the capacity of consumer loyalty programs to generate additional sales is essential for marketers who run such programs. However, customers' self-selection into the loyalty programs makes this evaluation difficult. This is the case especially in set-ups where the reward is not granted automatically upon achieving a certain number of points. In the case of automatic rewards, marketing theory predicts that points accumulation accelerates as consumers approach the threshold of necessary points for the reward, and is also boosted after the redemption, in what is called `the rewarded behavior effect'. In this thesis I use these insights to develop two models for evaluating loyalty programs where the rewards are not granted automatically. The first model applies to programs where consumers use the accumulated points like cash, for day-to-day expenses, while the second applies to programs where consumers use the points for non-ordinary treats, which on average are much larger. I estimate the parameters of both models using data provided by AIR MILES, Canada's largest coalition loyalty program. I show how sample heterogeneity and the non-random timing of the reward cash-in can be confounded with true loyalty program effects and I tease apart these effects to obtain non-biased estimates of program profitability. I use the model insights to suggest ways in which AIR MILES can change the program to further boost its profitability, contingent on retailers' contribution margins. The dissertation advances the literature by developing structural models for set-ups where retailers do not impose automatic redemptions upon consumers.
Nastasoiu, Mihaela Alina, "Evaluating Loyalty Programs with Endogenous Redemption" (2016). Electronic Thesis and Dissertation Repository. 3889.