A novel approach to study influencers' language in sponsored versus not sponsored posts and consumers' engagement

The widespread use of social networks completely changed the way of advertising products and services. An increasingly more used marketing strategy, known as influencer marketing, is to engage influencers for mentions on social networks, with the aim to maintain existing consumers, to increase them or to reach certain new ones. In this paper, we propose a novel approach to analyse the user generated contents on social networks produced by influencers for fashion brands, by also taking care of aspects related to the consumers’ engagement. The core part of the approach relates to the use of a suitable topic model, known as Structural Topic Modelling. This topic model allows to consider both the sponsorship statement and the number of likes received on Instagram posts, to also study the influencers’ language for sponsored vs not sponsored posts, as well as aspects related to the consumers’ engagement. We apply our proposal to Instagram posts produced by influencers, and related to two famous fashion brands, e.g. HM and Gucci, by also comparing the corresponding findings. The results are very satisfactory, confirming that the approach may be used by interested brand managers to efficiently define new marketing strategies, or to strengthen new ones.