Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

From Niche to Mainstream: Community Size and Engagement in Social Media Conversations

Jacopo Nudo, Matteo Cinelli, Andrea Baronchelli, Walter Quattrociocchi

TL;DR

This study addresses how platform architecture and community size shape public discourse across six platforms over 33 years. It employs a multi-platform longitudinal analysis using a two-dimensional participation density $D$ and a localization metric $L$, alongside crowd size $d$ and outreach $O_p(t)$ to quantify engagement and re-entry. Key findings show that smaller, niche platforms foster richer, longer interactions, while larger platforms enable broader but shallower participation; re-entry probability declines with increasing crowd and outreach, though niche platforms maintain engagement, with Dunbar-like limits observed in some contexts. The results offer practical guidance for platform design and policy aimed at fostering healthier, more meaningful online dialogue.

Abstract

The architecture of public discourse has been profoundly reshaped by social media platforms, which mediate interactions at an unprecedented scale and complexity. This study analyzes user behavior across six platforms over 33 years, exploring how the size of conversations and communities influences dialogue dynamics. Our findings reveal that smaller platforms foster richer, more sustained interactions, while larger platforms drive broader but shorter participation. Moreover, we observe that the propensity for users to re-engage in a conversation decreases as community size grows, with niche environments as a notable exception, where participation remains robust. These findings show an interdependence between platform architecture, user engagement, and community dynamics, shedding light on how digital ecosystems shape the structure and quality of public discourse.

From Niche to Mainstream: Community Size and Engagement in Social Media Conversations

TL;DR

This study addresses how platform architecture and community size shape public discourse across six platforms over 33 years. It employs a multi-platform longitudinal analysis using a two-dimensional participation density and a localization metric , alongside crowd size and outreach to quantify engagement and re-entry. Key findings show that smaller, niche platforms foster richer, longer interactions, while larger platforms enable broader but shallower participation; re-entry probability declines with increasing crowd and outreach, though niche platforms maintain engagement, with Dunbar-like limits observed in some contexts. The results offer practical guidance for platform design and policy aimed at fostering healthier, more meaningful online dialogue.

Abstract

The architecture of public discourse has been profoundly reshaped by social media platforms, which mediate interactions at an unprecedented scale and complexity. This study analyzes user behavior across six platforms over 33 years, exploring how the size of conversations and communities influences dialogue dynamics. Our findings reveal that smaller platforms foster richer, more sustained interactions, while larger platforms drive broader but shorter participation. Moreover, we observe that the propensity for users to re-engage in a conversation decreases as community size grows, with niche environments as a notable exception, where participation remains robust. These findings show an interdependence between platform architecture, user engagement, and community dynamics, shedding light on how digital ecosystems shape the structure and quality of public discourse.
Paper Structure (14 sections, 1 equation, 5 figures, 1 table)

This paper contains 14 sections, 1 equation, 5 figures, 1 table.

Table of Contents

  1. Introduction
  2. Related works
  3. Materials and methods
  4. Data Collection
  5. Preprocessing
  6. Measure of Users Participation
  7. Measure of User Re-entry
  8. Measures of Size
  9. Results and Discussions
  10. Distribution of Number of Users Across Platforms
  11. Analyzing User Participation
  12. User Re-Entry vs Crowd Size
  13. User Re-Entry vs Outreach Size
  14. Conclusions

Figures (5)

  • Figure 1: Log-log distribution of the number of unique users commenting under a post, across different platforms. The y-axis represent the absolute frequency of posts that have this amount of unique user. Note the heavy-tailed distributions typical of power-law behaviors.
  • Figure 2: Density matrices illustrating the distribution of a number of unique users ($d$), given thread prefix length ($k$, number of comments), across different platforms. The diagonal trend on platforms like Facebook and Twitter indicates a proportional increase in unique user engagement with thread length, while the lower concentration on platforms suggests a smaller core group driving most of the activity. Color intensity reflects the probability densities (conditioned by column), highlighting platform-specific interaction patterns.
  • Figure 3: Distribution of the localization parameter ($L$) across different platforms. The violin plot shows , using $L$, probabilities that an interaction (among a user and a post) is composed by a certain amount of comments, with each platform exhibiting distinct patterns. Localization values greater than 1 indicate a greater propensity to leave more than one comment in a conversation. While a value close to 1 stand for a distribution with less degrees of freedom.
  • Figure 4: Distribution of the localization parameter ($L$) across different platforms, controlling by the number of users involved in the thread (crowd size on x-axis). The number of users involved seems to influence the distribution of the number of comments posted by each user under the same post. Linearly for Usenet and Gab, with a saturation effect on Reddit and Voat, while it does not seem to impact giant mainstram platform such as Twitter and Facebook.
  • Figure 5: Plot of the localization parameter $L$ (probability of re-entry on y-axis) changing the outreach level of the community (outreach size on x-axis), across different platforms. Values on x-axis are represented in a Log scale for Facebook and Twitter. Each interaction has been grouped in a bin, based on the level of outreach that the community (or the page) hosting the interaction had at the time. It show how, almost everywhere the probability of re-entry in a conversation decrease with the increase of the outreach size, likely due to a progressively noisier conversation environment.