# Modeling Influencer Marketing Campaigns in Social Networks

**Authors:** Ronak Doshi, Ajay Ramesh Ranganathan, Shrisha Rao  
**arXiv:** [2106.01750](https://arxiv.org/abs/2106.01750)  
**Date:** 2021-06-03  
**Summary page:** https://www.summarizepaper.com/en/paper/2106.01750v3/

## Key points

- Social media platforms are widely used for quick and easy information sharing
- Influencers on these platforms have gained massive popularity and provide access to a vast potential customer base
- Selecting the right influencers for an advertising campaign is not straightforward
- The authors propose an agent-based model (ABM) to simulate influencer advertising campaigns
- The model incorporates real-world factors such as customers' interest, behavior, willingness to pay, brand's investment cap, influencers' engagement, and product nature
- Different types of influencers are evaluated based on customer acquisition cost and conversion ratio
- Results show that nano-influencers perform better than celebrities for luxury products but decline for non-luxury products
- Celebrities perform better as customers' interest decreases while nano-influencers' performance declines
- Influencer marketing strategies should be tailored to specific circumstances based on product nature and customers' interest levels
- Marketers can optimize their influencer advertising campaigns using this agent-based model

## Summary

In today's world, social media platforms are widely used for quick and easy information sharing, making them an effective tool for advertising products. Influencers on these platforms have gained massive popularity and provide access to a vast potential customer base. However, selecting the right influencers for an advertising campaign that can generate high returns with low investment is not a straightforward task. To address this challenge, the authors of this study propose an agent-based model (ABM) that simulates the dynamics of influencer advertising campaigns in various scenarios. This model aims to discover the best influencer marketing strategy by incorporating real-world factors such as customers' interest in a product, customer behavior, willingness to pay, a brand's investment cap, influencers' engagement with influence diffusion, and the nature of the product being advertised (luxury or non-luxury). Using customer acquisition cost and conversion ratio as unit economics, the performance of different types of influencers is evaluated under different circumstances simulated by varying the nature of the product and customers' interest. The results highlight the circumstance-dependent nature of influencer marketing and provide insights into which types of influencers would be more effective under specific circumstances. For example, when comparing luxury and non-luxury products, it is observed that nano-influencers show improvement while celebrities' performance declines. Additionally, as customers' interest decreases nano-influencers' performance declines while celebrities' performance improves. Overall, this research emphasizes that influencer marketing strategies should be tailored to specific circumstances and provides valuable guidance on selecting appropriate influencers based on product nature and customers' interest levels. By leveraging this agent-based model marketers can optimize their influencer advertising campaigns to achieve higher returns on investment. The findings presented in this study are published in IEEE Transactions on Computational Social Systems.