Share this
Election Social Media Bias: X/Twitter vs. Threads & Bluesky
by Henry Chapman on Nov 14, 2024 12:00:00 PM
Using Social Listening Data To Show Platform Bias
Throughout this election, we've noticed how strongly platforms are segmenting by political opinions. Elon Musk began this segmentation after he acquired X/Twitter, with more progressive users fleeing to Threads and Bluesky, two newer microblogging platforms that were all too happy to accept the new users. At Infegy, we've been collecting X/Twitter data since 2007, and we added Bluesky and Threads to our dataset earlier this year. Infegy's Social Dataset collects platform-wide, and our database is fast enough to give you a high-level overview of what audiences look like per platform (an analysis that no other social listening data provider can provide you). Let's dive into the types of people X/Twitter, Threads, and Bluesky attract after a polarizing election.
Our Query Structure For This Brief
For this type of broad-based analysis to get a general feel for how a platform reacts to a particular person (In this case, President-Elect Donald Trump), we rely on broad queries using the Donald Trump entity, which returns hundreds of millions of results. For each of our three queries (X/Twitter shown below), we'll add site filters so that we only see conversations on a particular platform.
We take this broad-based approach because if you get too specific, you will cut out a lot of meaningful perspectives from users. When we teach query writing, we encourage users to go as broad as possible because if you preemptively add too many filters, you'll exclude meaningful conversations and not even understand what you're excluding. In this case, we're trying to get a feel for the entire platform's perspective (on average); we'll return posts that mention Trump or other Trump-adjacent keywords. Other platforms can't do this, but Infegy Starscape can.
Figure 1: Query structure for the brief looking at general Donald Trump conversation over the last 3 weeks; Infegy Social Dataset.
Disparate Sentiment Across Platforms
We'll examine the different sentiment levels across our analyzed platforms for our first clue. We detected that X/Twitter had, on average, 8 percent less negative sentiment than Threads and 10% less negative sentiment than Bluesky. That means, on average, X/Twitter had much less negativity about Trump than the other platforms. While 8-10% less seems like little, our same dataset from X/Twitter included 14.4M records, an enormous amount of conversation.
Let's put that sentiment change into perspective. When our NLP algorithms classify content as positive or negative, we use context clues within the sentence. Much of the underlying subject matter tends to be highly negative when dealing with Trump-related content. The Trump Campaign frequently discussed immigration-related crimes, economic distress, and attacks against his political opponents. Therefore, this sentimental disparity across platforms and the tens of millions of underlying posts is worth looking into. To do that, let's dive into the personas we analyzed across these platforms.
Figure 2: Negative Sentiment Across X/Twitter, Threads, Bluesky (October 23, 2024 through November 13, 2024): Infegy Social Dataset.
Taking A Closer Look At Personas
So far, we've detected that Trump-related conversations on Twitter are 8-10% less negative than those on Threads and Bluesky. However, while helpful, average sentiment can only take you so far, as it abstracts much of the analytical context into just one figure.
To get a better perspective on what's thriving that disparate negative sentiment, we'll use Infegy AI's Personas feature. We can automatically generate personas using our underlying understanding of the accounts we collect from and a large language model to make the analysis human-readable. Let's dive into Infegy's AI-generated personas to figure out why.
Top X/Twitter Personas
Immediately upon running our AI personas, we find our answer. The top three personas we identified were all critical demographics for the Trump Campaign. Much of the Trump campaign's success relied on his religious, male base. Interestingly, we also detected a relatively new alliance that the Trump campaign has built, the cryptocurrency community. After building solid bridges with crypto, promising to remove cryptocurrency regulation, and appointing JD Vance as his vice presidential pick, a known cryptocurrency ally, it's no surprise much of the underlying positive conversation was driven by pro-crypto accounts.
Figure 3: Pro-Trump Conservative Activists Persona on X/Twitter (October 23, 2024 through November 13, 2024): Infegy Social Dataset.
Figure 4: Crypto and Financial Commentators Persona on X/Twitter (October 23, 2024 through November 13, 2024): Infegy Social Dataset.
Figure 5: Religious Conservative Voices on X/Twitter (October 23, 2024, through November 13, 2024): Infegy Social Dataset.
Leading Threads and Bluesky Personas
Let's repeat that same analysis for Threads and Bluesky. You'll note that we found progressives! Progressive Democratic Activists were the leading personas on both Threads and Bluesky platform-wide. We found that sample quote from Threads particularly telling when the profile says, "All maga get instantly blocked." It indicates the increased polarization on social media - specifically, a lack of willingness to discuss the complicated issues that make up the 2024 presidential campaign. There have been many discussions around political siloing based on news feeds. Here, we're seeing that division move a step further with platform-based political silos, removing any chance of positive discourse between opposing sides.
Figure 6: Progressive Democratic Activists on Threads (October 23, 2024, through November 13, 2024): Infegy Social Dataset.
Figure 7: Progressive Democratic Activists on Bluesky (October 23, 2024, through November 13, 2024): Infegy Social Dataset.
Takeaways For Your Brand
Social listening insights from Infegy highlight how election coverage is no longer just a matter of political affiliation but is now also a story of platform segmentation. With platforms like X/Twitter, Threads, and Bluesky showcasing distinct user biases, brands have an unprecedented opportunity to align messaging with the dominant sentiments and personas unique to each social channel.
For brands, this segmentation means that audience expectations and content performance can vary significantly by platform. Infegy's ability to analyze high-level personas across platforms can give your brand a tactical edge by matching message tone and positioning to platform-specific attitudes and community dynamics. For instance, conservative and cryptocurrency-enthused voices dominate X/Twitter's pro-Trump conversations, while Threads and Bluesky foster more progressive personas.
Adapting to these distinctions can help your brand craft platform-targeted campaigns, increase engagement with receptive audiences, and establish more nuanced and authentic conversations with your target demographics.
Key Insights: Election Media and Social Platform Dynamics
Platform Segmentation Impact: Social media platforms are becoming increasingly segmented by political opinions, with X/Twitter, Threads, and Bluesky attracting distinct audiences.
Sentiment Analysis: X/Twitter displays 8-10% less negative sentiment towards Trump than Threads and Bluesky, indicating platform-based sentiment variances.
Persona Insights: X/Twitter's prominent personas include conservative and cryptocurrency communities, while Threads and Bluesky lean towards progressive activists.
Brand Opportunities: Understanding platform-specific persona and sentiment can help brands tailor messages, drive engagement, and foster authentic interactions.
Strategic Approach: Leveraging Infegy's AI to craft campaigns aligned with platform-specific community dynamics can provide a competitive edge in campaign success.
Share this
- November 2024 (2)
- October 2024 (4)
- September 2024 (2)
- August 2024 (2)
- July 2024 (2)
- June 2024 (2)
- May 2024 (2)
- April 2024 (2)
- March 2024 (2)
- February 2024 (2)
- January 2024 (2)
- December 2023 (3)
- November 2023 (4)
- October 2023 (4)
- September 2023 (4)
- August 2023 (4)
- July 2023 (4)
- June 2023 (3)
- May 2023 (5)
- April 2023 (3)
- March 2023 (6)
- February 2023 (3)
- January 2023 (4)
- December 2022 (2)
- November 2022 (3)
- October 2022 (4)
- September 2022 (2)
- August 2022 (3)