Skip to content

MIDS-at-Duke/uds-2024-capybara

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

54 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Causal Effects of Virality on Baseline Per-Post Engagement Levels

In the rapidly evolving landscape of digital marketing, social media platforms stand out as critical arenas for business-to-consumer (B2C) engagement. Companies create social media profiles, not just for presence online, but to actively engage and expand their reach to a target audience. This digital engagement serves as a direct channel for disseminating a company's brand narrative, aiming to capture consumer attention in increasingly crowded markets.

This study focuses on Instagram post engagement within the fashion industry, seeking to understand how viral posts influence baseline engagement levels. Given the ephemeral nature of viral content, it is imperative to discern whether these spikes in engagement translate into sustained per-post engagement increases or if they merely represent temporary surges.

Using data provided by SocialInsider, our analysis delved into the per-post engagment trends for 20 fashion brands between March 2023 and March 2024. We employed a Difference-in-Differences (DiD) analysis isolating the effects of content virality, determining change in baseline engagement levels before and after a viral post. We performed a sensitivity analysis on the number of posts analyzed before and after the viral post to quantify the short term versus long term effects of virality on baseline engagement levels.

Our analysis shows that virality, characterized by extreme spikes in engagement, has a slight positive impact on social media baseline engagement rates when averaging, at minimum, 20 posts after the viral post. However, the practicality of this increase is not clear. When looking at the value of the increase (0.032 standard deviations of square root engagement) the data indicates that while viral posts do elevate visibility and interaction momentarily, they do not alter the baseline engagement levels. As for the slight reduction in engagement immediately following the viral post, we believe this is a byproduct of the viral post bringing heightened, transient traffic to posts immediately prior the viral post, thus artificially inflating the baseline just prior to the viral post. In effect, we are not seeing a reduction in baseline engagment, but rather a return to normalcy immediately following the viral post. This effect is lessened over time as the window beyond the viral post is expanded to 20 and 30 posts.

The strategic implications of these findings are significant. For brands, especially in the highly visual and dynamic fashion sector, our study suggests that relying solely on virality is not an effective method for gaining consistent content engagement growth. While viral posts can serve as brand awareness milestones for SEO strategy, they do not establish a durable foundation for ongoing audience engagement. Brands should aim for a balanced approach that cultivates a dedicated following through consistent, high-quality content rather than chasing the fleeting spikes of viral fame. This strategy not only enhances engagement but also builds a loyal community that supports sustained digital marketing success. ... read more

About

uds-2024-capybara created by GitHub Classroom

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •