Specifying the Treatment as the Repeated Effect
In a crossover trial, you can specify the treatment effect as the repeated effect. That is, since each type of coupon is sent to each customer, you can apply a covariance structure to the levels of Coupon value, using Customer ID to specify the subjects.