Satisfaction and loyalty are key determinants of word of mouth (e.g., Wangenheim and Bay on 2007b; Westbrook 1987). Because purchase amount and frequency often serve to operationalize loyalty (for an overview, see Verhoef 2003), we included customers' activity level (airtime) in our model. Innovators among the customer base are more influential and articulate more word of mouth (Engel, Kegerreis, and Blackwell 1969), so we also included MMS and HSDPA usage, which were in the beginning stages of their product life cycle in 2006 and mainly used by innovators. Wangenbeim and Bayón (2007b, p. 244) find that new cus-tomers of a firm are more likely to articulate recommendations because they "try to communicate the 'goodness' of their choice to others, either to convince themselves or to prevent others from disregarding their ability to make good choices." Therefore, we predicted a negative link between relationship length and CRP participation. Compared with nonparticipants, CRP participants might be more deal prone, because they are interested in the reward (Ryu and Feick 2007). Thus, we included status and status points eamed in the company's affinity program, with the assumption that customers interested in gaining rewards from the affinity program would also be interested in gaining rewards from the CRP.