Mega or Micro? Influencer Selection Using Follower Elasticity.

Posted on by

Reference:

Tian, Z., Dew, R., & Iyengar, R. (2024). Mega or Micro? Influencer Selection Using Follower Elasticity. Journal of Marketing Research (JMR), 61(3), 472–495. https://doi-org.devinci.idm.oclc.org/10.1177/00222437231210267

Keywords: Influencer marketing ,Streaming video & television, Causal inference, Deep learning Tags (Metadata) Online social networks

Summary: this article adresses the criteria for selecting influencers to partner .

While some firms collaborate with“mega” influencers with millions of followers, other firms partner with “micro” influencers with only several thousand followers,
but who also cost less to sponsor. To quantify this trade-off between popularity and cost, the authors develop a framework for
estimating the follower elasticity of impressions (FEI), which measures a video’s percentage gain in impressions (i.e., views) corresponding
to a percentage increase in the number of followers of its creator. Computing FEI involves estimating the causal effect
of an influencer’s popularity on the view counts of their videos, which is achieved through a combination of (1) a unique data set
collected from TikTok,

(2) a representation learning model for quantifying video content, and (3) a machine learning–based causal
inference method. The authors find that FEI is always positive, averaging .10, but often nonlinearly related to follower size.

They examine the factors that predict variation in these FEI curves and show how firms can use these results to better determine
influencer partnerships.

Development: 

Concepts :

Social network and influencer Marketing : how the structure of social networks can
impact downstream micro- and market-level outcomes.

  • Mega-influencers are useful for awareness.
  • Micro-influencers are perceived as more authentic and relatabl

Advertising:  the authors position influencer marketing as a form of social media advertising.

They study  the relationship between advertising investment and performance outcomes like sales or engagement.

The main research objective here is to know How does an influencer’s popularity (number of followers) causally affect the impressions/views of their videos?

in order to answer this  question the authors adobt a   causal inference and machine learning framework :

  • Representation Learning using Variational Autoencoder (VAE) to analyze video content.
  • Deep Instrumental Variables (Deep IV) to estimate the causal relationship between followers and impressions with content type and appeal type as moderators.

The data was collected from the Discover page of TikTok over six months (October 2020 – April 2021).

  • 216 hashtags.
  • 30 sponsored hashtags.
  • More than 500,000 videos.

Key findings:

FEI is positive but non linear meaning that the following elasticity of impressions is positive on average so more followers generally increase impressions but the relationship is not linear.

Midtier influencers generate the most marginal returns compared to mega or micro influencers.

the effectiveness of the influencer depends on the content type (category) and the engagement goal (informative, entertaining, socializing).

Limitations: 

-Tiktok specific content : the data was only collected through the tiktok platform and no other social media platforms so the results cannot be generalized all across platforms.

-The focus was only on impressions / views and didnt focus on sales or purchase behavior which is the ultimate goal.