We investigate experimentally the effects of priming decisions in an anti-coordination problem (i.e., El Farol Bar Game) with social expectations (i.e., empirical expectations and normative expectations). Specifically, we run online two connected experiments: in Experiment 1 we elicit incentivized social expectations and we make subjects play the El Farol Bar Game afterward; in Experiment 2 subjects play two different rounds of the El Farol Bar Game, receiving a prime with different social expectations from Study 1 - except for the first round of the Baseline treatment where no priming is implemented. We find that only empirical expectations drive decisions after the elicitation of the social expectations; while personal normative beliefs compensate for the uncertainty involving normative expectations. Priming decisions with information about social expectations only make people adopt more cautious behaviour (i.e. staying home), and not risk a higher payoff (i.e. going out). Further, a small fraction of subjects switch decisions when primed with a different social expectation in the second round, but the share of switching does not vary significantly across conditions.