This article explores the effects of time and relationship strength on the evolution of customer revenge and avoidance in online public complaining contexts. First, the authors examine whether online complainers hold a grudge-in terms of revenge and avoidance desires-over time. They find that time affects the two desires differently: Although revenge decreases over time, avoidance increases over time, indicating that customers indeed hold a grudge. Second, the authors examine the moderation effect of a strong relationship on how customers hold this grudge. They find that firms' best customers have the longest unfavorable reactions (i.e., a longitudinal lovebecomes-hate effect). Specifically, over time, the revenge of strong-relationship customers decreases more slowly and their avoidance increases more rapidly than that of weak-relationship customers. Third, the authors explore a solution to attenuate this damaging effect-namely, the firm offering an apology and compensation after the online complaint. Overall, they find that strong-relationship customers are more amenable to any level of recovery attempt. The authors test the first two issues with a longitudinal survey and the third issue with a follow-up experiment.

No abstract

The extant literature has studied the effects of a firm's service recovery efforts on the reactions of customers and employees following an individual service failure. However, the impact of recovery efforts on a firm's performance after a public and large service failure-such as a largescale information breach-has received scant attention. To address this gap, this current research develops a framework and finds support for the impact of service crisis recoveries on a firm's performance, as measured by firm-idiosyncratic risk. Using a unique dataset of service crisis recoveries, the authors find that firms offering compensation (i.e., tangible redresses) or process improvement (i.e., improvements in organizational processes) show more stable performance (less idiosyncratic risk), from two quarters to two calendar years after the announcement of their recovery plan. In line with the documented dual effect of apologies, firms that offer apology-based recoveries display more volatile performance (higher idiosyncratic risk). Of note, this volatility increases with the number of affected individuals, and it remains unaffected even when the apology is expressed with high intensity.

Although several meta-analyses have been conducted on the effectiveness of warning labels, many questions regarding their effectiveness remain unanswered. The authors identify 243 effect sizes from 66 primary articles, more than three times the number of effect sizes included in the most comprehensive meta-analysis to date. This updated and substantially larger data set shows that label effectiveness is contingent on the type of expected behavioral outcome. Labels aimed at moderation/cessation display a generally diminishing cascade of effects from attention (r = .32), comprehension (r = .37), recall (r = .31), judgment (r = .22), and behavior (r = .18). Labels targeting safe use show stronger effect sizes for behavior (r = .39) despite displaying a downward trend for attention (r = .35), comprehension (r = .29), recall (r = .32), and judgment (r = .21). The authors also find evidence of increased effectiveness when preactivating the label by means of an integrated communication strategy (r = .49). In addition, the results show the impact of several contextual factors (e.g., social influence [r = .33] and exposure frequency [r = .12]).

Decision-making, Involvement, Maximizers, Choice, Satisfaction, Retail,