Despite the explosive growth of influencer marketing, wherein companies sponsor social media personalities to promote their brands, there is little research to guide companies' selection of influencer partners. One common criterion is popularity: while some firms sponsor ``mega'' influencers with millions of followers, other firms partner with ``micro'' influencers, who may only have several thousands of followers, but may also cost less to sponsor. To quantify this trade-off between reach and cost, we develop a framework for estimating the follower elasticity of impressions, or FEI, which measures a video's percentage gain in impressions corresponding to a percentage increase in the follower size of its creator. Computing FEI involves estimating the causal effect of an influencer's popularity on the view counts of their videos, which we achieve through a combination of a unique dataset collected from TikTok, a representation learning model for quantifying video content, and a machine learning-based causal inference method. We find that FEI is always positive, but often nonlinearly related to follower size, suggesting different optimal sponsorship strategies than those observed in practice. We examine the factors that predict variation in these FEI curves, and show how firms can use these results to better determine influencer partnerships.

Keywords: influencer marketing, causal inference, deep learning, representation learning, heterogeneous treatment effects, video data

Influencer marketing, which involves a collaboration between firms and influencers, has shown to outperform other marketing strategies by earning consumers' trust and bringing higher conversion to firms. When cooperating with influencers, firms can leverage their established authority in a specific area to access a huge audience and affect its consideration and purchase decisions. However, influencers value their trustworthiness which they do not want to lose by writing inflated reviews for firms. In this paper, I try to reconcile the above tension by exploring 1) when a firm could benefit from cooperating with an honest influencer, 2) how the firm could design the informativeness of the influencer's message to maximize profit from the cooperation, and 3) how the optimal cooperating and information revelation decisions change by the firm type (established vs. new). We find that the firm can achieve positive profit from cooperating with the honest influencer even when its product is of fairly lower-than-expected quality since it can asymmetrically gain from the uncertainties introduced through the influencer’s review. The firm should further leverage those uncertainties by decreasing the informativeness of the influencer’s review when its product is of mediocre quality. Finally, the new firm benefits more from the less informative review and can generally achieve higher profit from the cooperation. Our results have managerial implications on when firms should strategically adopt influencer marketing and how they could design the information asymmetries and manipulate their product selection to maximize profits from influencer sponsorships.

Keywords: influencer marketing, information design, Bayesian persuasion, game theory

In this project, we investigate the unique form of influencer marketing via sponsored topics on TikTok (commonly found on Twitter's and Weibo's Trend page as well), which we call the "campaign hashtags", and explore sources of variations that drive their success on engaging users to browse (i.e. popularity) and participate (i.e. content variety). Specifically, a campaign hashtag is consisted of a name, a short description, and a set of videos under it created first by the sponsored influencers (i.e. the seeding videos) followed by the leading participants. We hypothesize that all three could be associated with the total amount of attention and organic content creation under a hashtag and develop machine learning algorithms to automatically 1) get embedding-based unexpectedness measure for each hashtag, 2) classify hashtags into substantive topics, 3) quantify the amount of imitation/innovation across the top videos, and 4) predict seeding video popularity and examine their ranking optimality. We find that the unexpectedness of a hashtag exerts a significant U shaped/inverted U-shaped effect on its popularity/content variety. Higher early content-creating conformity helps a hashtag become more popular, which is more prominent under socializing and narrower-scoped hashtags. For campaign hashtags specifically, ranking seeding videos based on their predicted performance has shown to significantly improve a campaign's aggregate popularity.

On social media, influencers might differ in their level of experience by which I classify them into the “professionals,” who are already established in their niches to get sponsorships from firms, and the “amateurs” who have yet to be sponsored. I then investigate the relationships and dynamics between these different experience levels. For example, in order to profit from their posts sooner, the amateurs might imitate the content of the professionals’ posts. Specifically, if a professional influencer posts a sponsored review for a product, then the amateurs might follow by reviewing the same product themselves with the hope that they might get higher exposure. For firms, these imitation reviews act as free advertising in addition to the paid sponsorships. To study these dynamics, I model the incentives behind imitating reviews, and use Instagram panel data to empirically evaluate the effectiveness and prominence of amateur imitation. In particular, I design imitating scores that quantify to what degree an influencer is using an imitating strategy and what imitating strategy is used, by comparing post similarity from different dimensions via machine learning methods. The results would shed light on whether firms should sponsor more influencers (if the amateurs imitate popular topics), more influential influencers (if they imitate successful idols), or influencers from different niches (if they imitate surprising products), and how these decisions differ by niche categories.

Firms want influencers to create content built around the preferences of their target audience. Analogous to product attributes that segment the market demand of a product across consumers, there should also be a set of underlying content attributes that relate to and differentiate browsing preferences across users on social media. Current tools (e.g., Facebook’s Creator Studio) have already suggested creators to test their video performance over some simple attributes (e.g., captions). However, given the high-dimensionality of videos, their attributes are endless and can be defined at different levels of abstraction. In addition, attributes potentially unknown to the researcher may significantly impact user preferences. Therefore, to find such attributes in a data-driven yet still semantically interpretable way, in my most recent project, I design an interpretable deep learning approach with a multi-level attention mechanism that automatically learns latent user preferences toward a set of latent video attributes, given the user information, ratings on videos, and processed video features. Attention fusion happens both at the demographics level (in the user network) and at the frame and modality level (in the video network) whose weights introduce interpretation to the learned preferences and video attributes. I am in the process of applying the approach to beauty videos (one of the most popular influencer niches), with the hope of discovering unobvious preference heterogeneity and cross-modality attributes useful for designing new videos beyond typical user groups (e.g., gender) and simple video features (e.g. captions).