When you see a social media ad with dozens of “likes,” do you assume it is worth your time? A new study suggests the answer is more complicated than it looks.

HKUST’s Professor Song Lin and his colleague conducted a large-scale field experiment on WeChat, analyzing more than 5.5 million ad impressions across 82 ads. Their goal was to find out whether showing more visible “likes” makes people more likely to engage with ads.

In the experiment, users were randomly shown ads with no likes, ads with one like, or ads with all the organic likes already gathered. The results showed that the first like had a powerful effect. Ads with a single visible like generated nearly one percentage point more likes and clicks than those with none—a significant increase given the typically low engagement rates.

Beyond that first like, however, the results split. More displayed likes consistently encouraged people to add their own likes, but they did not increase clicks. Users were happy to signal approval publicly but not necessarily to take the private step of clicking to learn more.

The researchers attribute this to two forms of social influence. Normative influence is the pressure to conform when others are watching, while informational influence is the tendency to treat others’ actions as evidence of value. Because likes are public, they trigger conformity. Clicks, by contrast, are private. When people suspect that likes are driven by social pressure rather than genuine enthusiasm, they stop treating them as a useful signal. This creates a conformity boost for likes but a crowding-out effect for clicks.

The study also revealed differences by brand and user type. For lesser-known brands, more likes did raise click-through rates, because users lacked other information. For well-known brands, extra likes added little. Likewise, socially active users were more swayed by visible likes than less engaged ones.

The findings were reinforced through three online lab studies. These showed that making likes visible strengthened conformity, while suggesting that likes reflected peer pressure reduced people’s willingness to click.

The lesson for advertisers is clear. Social proof can be a powerful tool for boosting brand visibility, but it does not always drive direct action. Popularity may look impressive, yet it does not guarantee clicks. Advertisers and platforms need to decide carefully when and how to highlight social cues.