By 2023 the global influencer marketing industry had grown to $21.1 billion (£15.5 billion) and many organisations are now increasingly turning to influencer accounts to help drive their sales.
But how companies assess the value of one influencer compared to another is difficult to ascertain. New research from UCL School of Management offers fresh insight into how companies should use influencers in markets shaped by online reviews and social learning.
Co-authored by Dr Mengzhenyu Zhang, the study looks at why influencers matter, how much value they create and which types of influencers firms should work with. It is one of the first papers to examine influencer marketing in organic market settings where customers tend to learn from each other by watching what others buy and reading their comments.
The researchers found that the effect of social learning is more subtle than many companies assume. It is often thought that the spread of organic reviews reduces the need for influencers, but the paper shows that the opposite can be true.
When customers already believe a product is likely to be good, an early endorsement from a trusted influencer can strengthen their confidence. This can make them less sensitive to later negative feedback, which helps sustain sales during a launch.
Speaking about the paper Dr Zhang said :
“Our findings show that social learning fundamentally changes how firms should think about influencer value. Contrary to the common intuition that social learning weakens the role of influencers, We show that social learning can increase the effectiveness of influencer endorsements when purchase intentions are moderate.”
The study also reveals that not all mistakes by influencers are equal. Endorsing a bad product can be more damaging than failing to praise a good one. Customers respond quickly to signs that an influencer has been too generous, which erodes trust in future recommendations. By contrast, a cautious influencer who occasionally withholds praise is still seen as credible and can create value for the firm.
When firms are choosing between influencers, accuracy matters even more when social learning is present. The paper shows that companies should favour influencers who very rarely endorse poor products. These influencers give customers a signal that stays robust even as more reviews emerge over time.
The authors also explore what happens when a firm has private knowledge about product quality. In these cases the firm cannot use an influencer to strategically shape the message. Any attempt to hide negative information would immediately reveal the firm’s intentions. As a result, firms are sometimes better off not knowing the product’s true quality before launch, since uncertainty allows them to commit to more effective information strategies.
“When firms can choose whom to hire, the optimal strategy favours positively biased information—endorsing good products with certainty while occasionally endorsing bad ones. Importantly, in the presence of social learning, the most effective influencer must be more informational, with a lower tendency to issue false positives, so that her endorsement conveys stronger informational content.”
