Evaluating Twitch's Role as a Digital Third Place

ROLE

Content Analyst

YEAR

2024 - 2025

TLDR

With the concept of digital third places becoming increasingly popular, I took a deep dive into Ray Oldenburg's original characteristics for third places, Charles D Soukup's digital adaptation of these characteristics and then performed my own analysis to uncover whether Twitch really earns its spot as a digital third place in today’s online world.

What

The term "digital third place" is often used to describe online communities or virtual spaces where people gather to discuss interests, interact and build relationships, building on Ray Oldenburg's concept of (physical) third places. Twitch in particular is a live streaming platform that is frequently touted as a digital third place on the basis of its interactive live chat, real-time communication, and community-driven channels.

Why

There has been a lack of research that builds on empirical evidence (over conceptual arguments) to support this claim.

How

I used two content analysis methods to gather data and build my claim. I performed grounded thematic analysis on 1000 data units extracted from live stream chats to build themes without any preconceived notions. I performed feature analysis of Twitch's webpages to understand if there are any features that encourage or inhibit third-place interactional dynamics.

And then?

Based on my analysis, I concluded that while Twitch seems to support some aspects of digital third places, it's alignment with the deeper theoretical framework is superficial. Currently, I'm working on furthering this research to understand streamer-watcher relationships on Twitch.

RECORD SCRATCH

HOW DID I GET THERE?

My Research Questions

Based on my literature review of existing work in similar topics, I defined two research questions to direct my analysis.

RQ 1

What affordances do Twitch features provide that liken the platform to a digital third place?

Is this phenomenon of Twitch being called a digital third place supported/promoted by the platform as well? How/How not?

RQ2

Do features like subscriptions, bits, and donations create an unintentional social hierarchy in community interactions?

Lack of social hierarchy is a major characteristic for physical third places. How does it translate to digital third places?

Data Collection

Data for Grounded Theme Analysis

Each chat message was counted as a unit of data. 1000 cumulative consecutive messages were chosen from two livestreams. The total word count came up to 6140 words.

Data for Feature Analysis

Features were identified by careful observation of Twitch webpages. Qualitative data for feature analysis was gathered through observations on Twitch, user discussions on Reddit, and personal experiences with the platform.

Methods of Analysis

This study employs two key methods - Grounded Theme Analysis and Feature Analysis.

Grounded Thematic Analysis

I used grounded thematic analysis to identify and build themes for interaction dynamics in the live stream chats.

To analyze chat interactions, I used a modified grounded theory approach with two rounds of open and axial coding over 200 messages. This helped identify emerging patterns and themes in how viewers engaged with streamers and each other. The final codebook focuses on key interaction dynamics like casual communication, social status, recurring participation, and information passing. I also categorized the messages into two types: general chats and automated chats.

Table: Codebook for general chat messages

Table: Codebook for automated chat messages

For accuracy and context, the data was coded live - as in, coded while watching the live stream. This approach ensured greater precision in capturing real-time interactions and nuances.

Feature Analysis

I conducted a feature analysis of Twitch live stream pages to explore how platform design influences third-place interaction dynamics. I focused on features surrounding monetization such as subscriptions, bits, and donations to assess whether they act as social levelers or create hierarchies. I also focused on chat affordances like emotes, moderation tools, slow-mode, and follower-only chat to understand how they influence inclusivity and participation.

Results

Here, I report and discuss my most impactful findings.

Grounded Thematic Analysis

93.1% of the 1000 coded messages were general chats, indicating that the viewers largely featured in the chat spaces. Of the automated chats, a cumulative of 59.42% chats showed proof of some sort of monetization (donations, subscriptions). This presence of proof of monetization pulls chat spaces away from being neutral ground, creating an unintentional (and often subtle) social hierarchy.

Observation of chats in real time showed that while viewers could potentially start conversations and converse amongst themselves without the the streamer's input, the streamer was still the main driving force behind most if not all conversations. The streamer having to act as a host removes the opportunity for organic interactions, a vital part of third places.

50.59% of the general chat messages were classified as blurbs - messages that couldn't start or sustain conversations in any manner. While these blurbs often add on to the playful and low profile nature of these live stream chats, the short length and the rapid nature of blurbs tend to limit deeper exchanges from happening. This once again pulls Twitch away from a vital third place characteristic of "Conversation as main activity".

Feature Analysis

Live streaming and chatting is arguably the main feature of Twitch. It facilitates real-time interaction, but also the size of a streaming community significantly affects the quality and quantity of interactions that occur.

Features like Videos On Demands that provide asynchronous participations tend to go unnoticed, lacking the community building aspects that the other features provide.

Features like Subscriptions are vital for streamers, as this is one of their main sources of monetary compensation. However, they (often unintentionally) can create a social hierarchy, moving from the neutral ground that third places seek to provide.

Conclusion

The findings of the analysis by theme analysis and feature analysis suggest that Twitch features do not offer many affordances that align it with the concept of a digital third place. On a high level view, it offers a means to foster informal interactions, enable community building through live chat, and provide spaces for shared interests and social bonding. But upon observing the deeper theoretical framework laid out by Ray Oldenburg and Charles D Soukup, including acknowledgement of the nuanced understanding of unequal representation (unintentional via rapid messages from some viewers or intentional via follower-only and sub-only streams), western centric digital spaces and the potential for digital biases in the way Twitch communities are structured and interact, this alignment remains high-level.