One in five off-page citations in AI search responses comes from Reddit, and that share is growing 30% year over year.
This data point led to an entire session from Amanda Kusner, Senior Solutions Consultant at Uberall, and Peter Wischmann, Senior Client Partner at Reddit. Their main point: AI models decide which companies to recommend based on signals that most multi-site brands never monitor, and Reddit is at the top of that list.
Kusner and Wischmann combed through the entire citation data behind today’s real AI ranking factors, then laid out a five-part stack for multi-location visibility it starts with location data and ends with an operational rhythm that competitors struggle to copy. They also showed What a Live AI Citation Dashboard Reveals This is the first time that a brand has offered one.
Half of searches now end without a click. The question is whether the AI mentions your locations in the response.
Watch the full session on demand.
Why AI Search cites Reddit more than any other source
Why does AI search cite Reddit so often? Because AI systems create responses from human context, and Reddit delivers them at scale: around 500 million weekly active users in over 100,000 active communities.
When someone asks ChatGPT, Gemini, or Perplexity a question about your category, the model reads 5 to 16 different sources before responding. Your own website makes up about 15% of what it finds. The rest comes from Reddit threads, review platforms, directories and forums.
Within this off-page mix, Reddit leads all sectors tracked by Uberall. Wischmann clearly expressed the strategic consequence: “The transition is no longer from a ranking to a better ranking, but from a ranking to a recommendation. »
Kusner added that about three-quarters of companies are missing from AI conversations regarding their category. If a Reddit thread in your category exists and you’re not there, a competitor or wrong answer is filling the space.
The session showed exactly which invites and subreddits generate quotes in a given categoryincluding the gap mapping exercise that both speakers use with brands.
Google integrates Reddit threads into business profiles
Google has started making Reddit threads appear directly on business profiles, alongside the address, phone number, and hours.
This changes the stakes for each local thread. A question like “is this place worth it” in a local subreddit now appears to everyone searching for the business, whether anyone has answered it or not.
“If a customer asks a question in public, it’s not just noise, it’s a signal,” Wischmann said. “If this question remains unanswered, the brand leaves the possibility for a competitor or the Internet to answer on its behalf.”
The subreddits that Google brings up around your locations are not random. It is often the communities that already shape local consideration.
Action item: Search for your top locations on Google and note the threads that appear, then learn to identify the subreddits that shape your local reputation.
Your location data decides whether the AI mentions you or not
How does location data affect AI visibility? When your name, address, phone number, and hours conflict across Google, Apple, Yelp, and vertical directories, the template provides two options.
“Instead of determining which one is the right one, the AI will either choose one, and it might be the wrong one, or it will completely ignore you,” Kusner said.
The risk worsens on Reddit. A thread showing the wrong address or outdated hours can become the response that an AI model confidently repeats over and over again.
Kusner positioned clean location data as the basis of the session’s five-part stack: Local landing pages with schema markup and local FAQs provide the AI with quotable sources that come directly from the business, and reviews validate those pages’ claims.
The order of the five pieces is important because each builds on the previous one. View the full stack and its build order in the session.
How to appear on Reddit without being criticized
Reddit rewards participation and punishes promotion. Wischmann referenced an analogy with Reddit’s CEO: Reddit is a city, while other platforms are a stage.
A brand that reads the room, contributes something useful, and thanks the people who recommend it, creating public evidence of its relevance. A brand that offers a fancy banner-style design is dismissed, flagged, or held up as an example of what to avoid.
“AI doesn’t just reward the most optimized brand, it rewards the most credible,” Wischmann said. Credibility comes from places where people compare, validate, recommend and challenge in public.
The approach works at scale. Wischmann highlighted Carl’s Jr., which delivered a 176% behavioral improvement and 85% lower cost per visit compared to the benchmark, measured using Foursquare traffic data.
The session covers the commitment manual behind these resultsincluding when a branded subreddit makes sense and the mix of formats executed by Carl’s Jr.
Q&A: Most useful questions from the webinar
Q: What is the quickest win if I have limited time and resources?
Amanda replied: Start with an audit of your top 10-20 placements in the top directories, including Google, Apple, Yelp, and any vertical-specific listings. Confirm that the name, address and telephone number are accurate and consistent throughout. “This alone is going to close a gap that I think most brands don’t realize,” she said. From there, provide the AI models with questions your customers would ask and study what is cited. If nothing comes out of your website, this gap defines your next priority.
Q: Does responding to reviews actually count towards AI visibility?
Amanda replied: Yes. When your website claims fast service and friendly staff, and 200 reviews on Google and Yelp say the same thing, “AI can connect these dots” and treat that as validation. Unanswered reviews, especially negative ones, send the opposite signal: a brand that doesn’t pay attention to them. Responding creates the layer of trust that AI models look for when deciding who to recommend.
Q: We are a franchise model. How do we achieve alignment when we don’t control what franchisees do?
Amanda replied: Share the property. The company owns the infrastructure: list management, source of truth, and brand safety guardrails. Franchisees own the local context: community engagement, local events, and market-specific FAQs overlaid on the business foundation. She called this the most common challenge Uberall expects to face in multi-location businesses, and the split allows the companies to maintain visibility without removing local authenticity.
Q: For multi-location brands, should we start nationally, locally, or both?
Peter replied: Both, in the right order. Start with a broader national or regional layer so campaigns have enough scale and signal, then layer on local targeting for priority markets with stronger store density or business needs. Brands that start too early limit their audience before the platform can learn. Local targeting on Reddit works best in combination with community and contextual targeting.
Watch the full webinar
The full session includes the full five-part stack in build order, Geo Studio’s live dashboard walkthrough showing where the AI pulls its answers, the prompt mapping exercise, and the Carl’s Jr. campaign breakdown. Watch it on demand.
