Can attempts to delight customers with surprise gains boomerang? A test using low-price guarantees

There are apparent similarities between the context we examine and service failure/recovery (Dong et al. 2008; De Matos et al. 2007; Roggeveen et al. 2012). Thus, locating a lower price after purchase under an LPG is akin to service failure and the retailers’ subsequent offering of a refund (with or without an additional surprise component) is akin to service recovery. However, there is an important difference between our context and that involving service failure/recovery per se. Service failure can occur without a pre-purchase guarantee preceding a service, just as consumers might locate lower prices following a purchase that was not made under an LPG. However, our paper is about failure of a pre-purchase signal, and not the failure related to the purchase per se. Thus, locating lower prices when a purchase was not made under an LPG would not be considered a failure in the specific context we examine, whereas perceptions of service failure might occur even without a pre-purchase guarantee. The strict analogue of our situation in the service context is failure of service (and subsequent recovery efforts) when the service is qualified with a service satisfaction guarantee; which has not been examined in the service failure literature, to the best of our knowledge.

In this paper, we focus more on when offering surprise refund components backfire, and so reduce future purchase intention. Nevertheless, in some cases, offering surprise refund components may indeed increase delight, and so increase purchase intention. Given that this latter point is not the focus of this paper, we only briefly examine this issue (in a follow-up study to Study 2), examining how exactly delight increases future purchase intention.

Including these six respondents in the analyses did not materially change the results reported. For example, the interaction result reported in the next paragraph (F_{1, 84} = 19.61; p < .001) was materially unchanged when these six respondents were included (F_{1, 90} = 15.95; p < .001).

In Studies 1–2, based on pretests described in Appendix 3, we used a moderate signal default magnitude of $60. In Study 3, we had to choose a signal default magnitude smaller than $60, and yet one for which participants would exhibit refund-seeking behavior. Based on the pretests, we chose to use $30 as the small signal default magnitude.

Locating lower prices may be easier on the internet; however, some retailers explicitly exclude online-only prices when considering LPG default (Tuttle 2013)