ChatGPT Is Asking for Reddit by Name Before It Answers

Welcome back to What Matters This Week.

Last week, we shared data on the extent of Yelp’s dominance as an AI visibility factor in local business discovery. This week, new data from Profound reveals the upstream influence that Reddit has over one of the most popular AI tools in the world. 

ChatGPT Is Asking for Reddit by Name. 24x More Often Than Before.

Here’s the TL;DR

• New analysis from Josh Blyskal at Profound shows ChatGPT is writing “reddit” into its query fanouts 24x more often than in January. Reddit has returned to #1 in ChatGPT citations with an 8.5% share. 
• Fanouts are an upstream GEO signal. The model declares what kind of source it trusts before it cites anyone, meaning visibility is determined one layer earlier than most marketers measure.
• Pull fanouts for your top priority prompts. See where the model is reaching. If Reddit is one of those channels (and for most B2B categories, it now is), audit what’s being said about your brand there and build real presence. 

What’s Happening

In a LinkedIn post this week, Josh Blyskal shared LLM response data showing Reddit has reclaimed the #1 spot in ChatGPT citations, and that the model is now searching for Reddit content far more frequently than at the start of the year.

The analysis (co-authored with Brandon Punturo) draws on two datasets: 

1. Profound’s internal data tracking millions of LLM responses per day
2. Roughly 7 million real ChatGPT prompts from actual searchers between January and late May

Here is what the data shows:

• In January, roughly 0.15% of ChatGPT responses included fanouts mentioning “reddit.”
• By late May, that figure had climbed to 3.68%, a 24x increase in ChatGPT explicitly searching for Reddit by name.
• Reddit is now the #1 cited domain in ChatGPT, accounting for 8.5% of all citations.
• The spike in Reddit fanouts and the spike in Reddit citations move in lockstep.

The Reddit data is the headline. What sits underneath it is more interesting. ChatGPT is now explicitly deciding it wants Reddit content before it picks any source at all. That decision shows up in the query fan-outs, which until recently were invisible to most measurement tools. 

Why It Matters

Fanouts change how you measure AI visibility.

Citation data tells you who won the answer. Fanout data tells you which sources the model was willing to consult before it picked anyone. That’s a new analytical layer rich with source data that marketing teams need to monitor and measure.

Each source word the model adds to a fanout is a signal. When ChatGPT writes “reddit,”as the Profound data shows,  it’s signalling that the question requires lived human experience. When it writes “G2” or “Capterra,” it wants third-party review data. When “GitHub” or “docs” appear, it wants implementation proof. When “pricing,” “alternatives,” or “vs” appear, it wants buying-context pages.

Read across hundreds of prompts in your category, and you have a map of where the model has already decided your visibility lives. 

Reddit is currently the loudest fanout signal.

Of all the query modifiers ChatGPT could be writing into its fanouts, “reddit” is the one rising fastest. The 24x increase since January is far steeper than any other named source we have public data on, and it’s happening at the same time Reddit’s overall citation share has climbed to 8.5%, putting it back at #1.

The combination matters. Citation share alone could be explained by Reddit content getting indexed more aggressively. The fanout spike means something stronger: ChatGPT is now structurally deciding that Reddit input belongs in more answers, before it has even looked at any specific page.

For most B2B categories, that has a clear content implication. Another 1,500-word SEO blog post on your category will not change the outcome if the model has already decided it wants community input on that question. 

What to Do About It

• Audit query fanouts for your top 15–20 priority prompts. Tools like Profound, Writesonic, and AirOps now make this visible. Look at the sub-queries ChatGPT writes before it answers. The source words it adds reveal what kinds of content it trusts for those queries.
• Bucket the source words by type. Communities, review sites, technical documentation, comparison pages. This becomes your category-specific map of which channels actually matter for AI visibility in your space, rather than a generic “be everywhere” answer.
• Audit what Reddit currently says about your brand. For most B2B categories, Reddit will be among the channels ChatGPT is reaching for. If the threads ranking for your category contain outdated or hostile content about your product, that has moved from “marketing sidequest” to “GEO priority.” Use the Lurk, Listen, Leap approach to participate authentically.

🎙 Upcoming Event: What Still Works? Content, Search, and Distribution in the AI Era

If this week’s fanout data made you want a clearer picture of what actually wins AI citations, this one is for you. Ross is sitting down with Samanyou Garg, Founder and CEO of Writesonic, for a live masterclass on what’s driving AI citations across thousands of brands. 

Thursday, June 18, 1:30 PM ET. 60 minutes live on Zoom with audience Q&A.

The session covers what to stop doing in content and SEO in 2026, and where Reddit, community content, and third-party mentions fit in the citation economy. The promise: leave with something you can ship Monday morning.

👉  Save your seat for the free masterclass

Go Deeper on AI Visibility and Reddit

→ Research: Reddit Accounts for 21% of AI Citations for Key B2B SaaS Prompts. Our full study with AirOps mapping how 50 B2B brands across seven verticals appear in AI responses. Covers how citation patterns shift between branded and unbranded queries and which third-party sources dominate each vertical. [Foundation Labs]

→ Josh Blyskal’s full LinkedIn post on fanout features. Worth reading in full for his framing of fanouts as a new AEO signal, plus the source-word framework for interpreting them. [LinkedIn]

→ How Query Fan-Out Works, Why It Matters, and What It Means for Brand Visibility in AI Search. Our full breakdown of the eight types of fan-out queries Google’s system generates and what the AirOps research on 43,000+ ChatGPT searches tells us about how to structure content for AI visibility. The technical companion to this week’s issue. [Foundation Labs] 

That’s it for this week.

If something landed, tell us. If something felt off, tell us that too. Reply to this email or DM me on LinkedIn.

Have a great weekend,

Ethan Crump 

ethan@foundationinc.co