The current research questions if service recovery has differential effects on complainers depending on the way that they initially complain, being privately (e.g., emails, phone call) or publicly (e.g., social media, blogs). Using four studies, the current research offers several core contributions. First, building on justice theory, our findings show that a recovery is especially effective at appeasing private complainers' negative affect, while this same recovery has less impact for public complainers. Second, we show that for public complainers, the role of a recovery will be different depending on the level of public exposure. When public complaints are viewed by just a few observers on social media (i.e., low exposure), such complainers assess their own actions of justice restoration as being ineffective. Third, we find that our previous findings are robust no matter if the customer is a complainer or an observer.