Influencer Cartels (2405.10231v2)

Abstract: Social media influencers account for a growing share of marketing worldwide. We demonstrate the existence of a novel form of market failure in this advertising market: influencer cartels, where groups of influencers collude to increase their advertising revenue by inflating their engagement. Our theoretical model shows that influencer cartels can improve consumer welfare if they expand social media engagement to the target audience, or reduce welfare if they divert engagement to less relevant audiences. Drawing on the model's insights, we empirically examine influencer cartels using novel datasets and machine learning tools, and derive policy implications.

• The paper demonstrates that influencer cartels collude to artificially boost engagement, distorting advertising value.
• Data from Instagram and Telegram reveal that topic-specific cartels deliver 60-85% of natural engagement quality, while general cartels perform at only 3-18% value.
• The study recommends tightening regulations on general cartels and revising payment models to improve transparency in influencer marketing.

Understanding Influencer Cartels

Influencer marketing is a booming industry, with billions of dollars flowing towards influencers who promote products on social media. But what happens when influencers band together to artificially inflate their engagement metrics, and how does this impact advertisers and consumers? Let's delve into the intriguing world of influencer cartels, as explored in this research.

What's an Influencer Cartel?

In a nutshell, an influencer cartel is when a group of influencers colludes to manipulate their engagement metrics—likes, comments, shares—to appear more influential than they actually are. This is done to attract higher advertising fees. These cartels often operate through online platforms like Telegram chat rooms, where influencers agree to engage with each other’s content according to specific rules.

Here’s how it works:

• Influencers are required to like and comment on a set number of posts from other cartel members.
• In return, their own posts receive similar engagement from other members.
• An algorithm ensures that everyone follows the rules to get their "earned" engagement.

Why Influencer Cartels Form

One major reason for the formation of these cartels is the payment models in influencer marketing. Many influencers are paid based on past engagement metrics rather than the actual success of ad campaigns they run. This creates a strong incentive to game the system by artificially inflating engagement.

Impacts on the Market: The Good and the Bad

The paper's theoretical model provides intriguing insights:

• Positive Externalities: A small, topic-specific cartel can potentially improve overall engagement quality. If influencers with closely related topics promote each other's content, it could benefit both the infuencers and their audiences.
• Negative Externalities: General cartels, where influencers engage with each other's content regardless of topic relevance, can flood social media with low-quality engagement. This misleads advertisers and harms consumers by showing them less relevant content.

Influencer Cartel Types: General vs. Topic-Specific

The paper empirically validates its theoretical model using data from both Instagram and Telegram. Here are the key findings:

• General Cartels: These cartels, which do not restrict engagement to specific topics, generate significantly lower-quality engagement. The engagement quality from these cartels is nearly as poor as random engagement.
• Topic-Specific Cartels: These cartels, focusing on narrow topics like "fitness" or "fashion," generate higher-quality engagement, closer to natural interactions.

Key Metrics and Analysis

To measure engagement quality, the researchers use advanced machine learning models to quantify the similarity between the content of influencers and their engager:

• They use the LaBSE model for text-based analysis and the CLIP model for combined text and photo analysis.
• Cosine similarity measures how closely related the content of the engager is to the influencer's content.

Numerical Insights

The research translates these theoretical findings into practical implications with some insightful calculations:

• For general cartels, advertisers get only 3-18% of the engagement value they pay for, relative to natural engagement.
• For topic-specific cartels, advertisers still overpay but get 60-85% of the engagement value.

Policy Implications

Based on these findings, the researchers suggest several policy directions:

1. Regulate General Cartels: Given their detrimental impact, shutting down general cartels is likely beneficial.
2. Expand Current Regulations: Regulatory frameworks should also address in-kind exchanges of fake engagement.
3. Change Compensation Models: Moving away from payment models based solely on past engagement metrics to those based on actual campaign success could deter such collusive behaviors.

Future Perspectives

The paper's insights point towards a future where influencer marketing could become more transparent and efficient with proper regulation and improved compensation models. While topic-specific cartels might offer some benign cooperation, general cartels clearly produce more harm than good.

Final Thoughts

Understanding and addressing the dynamics of influencer cartels is crucial as the influencer marketing industry continues to grow. By ensuring that engagement is genuine and valuable, advertisers can make better-informed decisions, and consumers can enjoy more relevant content. This research provides a critical step towards achieving that goal.