Purpose: This paper aims to investigate if product pre-announcement effects measured using stock market returns conform to the predictions of two competing consumer marketing theories. In particular, while buzz marketing theory indicates a direct positive effect, information asymmetry theory suggests an influence contingent upon evidence. The study also investigates whether a pecking order of performance effects exists across different signaling situations. Design/methodology/approach: The final sample consists of 219 product-preannouncements reported in the Wall Street Journal between 2005 and 2015. The standard event study methodology was used to test for performance effects. Findings: The results show that preannouncements with evidence alone significantly outperform those with buzz alone, and announcements containing buzz and evidence. Also, buzz acts as a salient moderator of the relationship between evidence and performance. In addition, company size also affects the evidence-performance relationship, with smaller firms benefiting more from evidence than larger firms. Research limitations/implications: The event study method assumes efficient markets and deals with publicly traded companies. Practical implications: Managers can allocate resources wisely by deciding whether to invest in evidence or buzz in their pre-announcements. Originality/value: In contrast to extant research that primarily investigates contingency effects, this study identifies how an important moderator, i.e. buzz affects performance.