- ReddVisible
- Posts
- Why AI Skips Most Brands
Why AI Skips Most Brands
Reddit and location data now determine who AI recommends, and most businesses aren't even in the conversation.
Ewen Finser
July 24, 2026
Welcome to this week’s edition of ReddVisible.
AI doesn't recommend the businesses that rank best. It recommends the ones it understands.
This week: why 75% of multi-location brands never appear in AI recommendations, how bad location data causes AI to skip businesses, and why Reddit has become part of local discovery infrastructure.
See what you missed from the last edition:
Let’s get started.
Reddit Now Powers One in Five AI Search Citations
One in five AI search citations now comes from Reddit, and that share is growing 30% a year. Yet 75% of businesses never appear in those conversations.
Context
Reddit already supplies one in every five off-page citations that show up in AI search answers, and that share is growing 30% year over year.
The mechanics explain why.
Reddit carries roughly 500 million weekly active users across more than 100,000 active communities, and AI models read five to sixteen different sources before they answer a question.
Your own website accounts for about 15% of what a model finds. The rest comes from Reddit threads, review platforms, directories, and forums.
To put it simply: success is no longer about ranking higher. It's about becoming the recommendation.
That is not a search-optimization problem. It is a visibility problem, and the number makes it concrete: roughly three-quarters of businesses are completely absent from the AI conversations already happening about their category.
Analysis
Google has started pulling Reddit threads directly onto business profiles, next to the address, phone number, and hours.
A single unanswered "is this place worth it" post in a local subreddit now surfaces to every searcher who looks the business up, whether or not the brand ever responded.
Location data quality decides what happens next.
When a business's name, address, phone number, and hours conflict across Google, Apple, Yelp, and vertical directories, the model has two options: guess which version is correct, or skip the business entirely.
When AI encounters conflicting information, it has two likely outcomes: recommend the wrong business or skip yours entirely.
A Reddit thread flagging a wrong address or outdated hours can become the answer a model repeats with total confidence, indefinitely.
This is the part most marketing teams miss.
They treat Reddit as an earned-media channel to monitor occasionally. The data says it now functions as core location infrastructure, sitting next to your Google Business Profile in the model's decision tree, not below it.
The scale gap compounds the problem.
Reddit's own communities are large enough that a single relevant subreddit can carry more decision-shaping conversation than a brand's entire owned content library, and AI models are already treating it that way when they choose which five to sixteen sources to cite.
Implications
The upside is measurable, not theoretical.
Carl's Jr. drove a 176% behavioral lift and an 85% lower cost per visit versus benchmark, tracked through Foursquare foot-traffic data, after treating Reddit engagement as part of its visibility strategy rather than a side project.
The playbook starts before any posting happens.
The fastest win is to audit your top 10 to 20 locations across every directory that matters, confirm your name, address, and phone number match everywhere, then prompt the AI models your customers would use and study what gets cited. If nothing pulls from your own site, that gap becomes the next priority.
The engagement half is where most brands lose the thread.
Think of Reddit as a city, not a stage.
A brand that reads the room, contributes something useful, and thanks people who recommend it builds public evidence of relevance. A brand that shows up with polished, banner-style creative gets downvoted, flagged, or held up as the example to avoid.
AI doesn't just reward the most optimized brand. It rewards the most believable one.
For franchise and multi-location brands specifically, that means splitting ownership: corporate owns the directory infrastructure and brand-safety guardrails, while local teams own the community engagement and market-specific detail that AI models are already reading as signal.
The math is simple: a service business that closes even a fraction of the 75% visibility gap through clean directory data and active subreddit participation is competing for citations that most of its category isn't even trying to win yet.
I'm Watching: I'm watching how many multi-location brands still treat this as a project to schedule for next quarter instead of a standing weekly habit. A 75% absence rate isn't a data gap. It's a strategy gap, and Reddit is the fastest place to close it.
Tap Into Reddit Paid Marketing With ScalePaid
After two years of running organic Reddit campaigns for brands, I kept seeing the same pattern. The posts that moved the needle were never the ones we engineered. They were the ones rooted in authentic engagement.
That insight led me to build something new.
For brands that have been curious about Reddit ads but have not had a specialist in their corner, ScalePaid fills that gap.
I partnered with Ben Dankiw to build ScalePaid, a dedicated Reddit paid ads agency.
Ben brings years of performance marketing expertise to what has been, until now, a purely organic operation.
Together, we are closing the loop between what works organically on Reddit and what deserves a paid amplification budget.
🎯Reddit's Own Ad Team Can't Agree If AI Deals Hurt the Business
Reddit executives are openly debating a question they spent years avoiding: whether licensing its community data to train AI models ultimately weakens the competitive advantage behind its own advertising business.
That uncertainty matters because Reddit sells advertisers on its ability to understand contextual intent better than other platforms through products like Reddit Max and Reddit Community Intelligence.
Those same conversations also help train external AI systems.
The distinction is that advertisers buy engagement signals, not raw text. Reddit licenses posts and comments for model training, while its ad products rely on how people interact with that content. Whether that separation proves durable remains an open question.
The broader paradox is becoming harder to ignore.
Reddit's content increasingly powers external AI systems even as Reddit invests heavily in AI to strengthen its own products.
The bigger strategic question isn't whether AI search reduces Reddit traffic. It's whether the data Reddit sells to train those models gradually erodes the unique advantage its advertising business depends on.
Interest-based targeting may soften that risk. Because Reddit organizes advertising around communities and intent rather than personal identity, it may benefit from AI-driven referral traffic even as identity-based platforms face greater disruption.
Practical Angle: If your media plan assumes Reddit's value is stable, it probably isn't. Reddit's own executives are still debating how its AI licensing business could reshape its advertising business.
Ask your Reddit rep directly how targeting signal is segmented from licensed training data, and whether that separation shows up anywhere in your campaign reporting.
Budget review cadence matters here: treat this as a quarterly question, not an annual one, since the "fundamental difference" Reddit is describing is still being defined in real time, by Reddit's own account.
Watch List:
Reddit's next quarterly filing for new risk-disclosure language on AI-driven traffic or ad-revenue cannibalization
Whether Reddit Max reporting starts separating logged-in engagement from AI-referred, non-logged-in traffic
Any changes to Reddit's licensing terms with OpenAI or Google that restrict using licensed data for advertiser targeting
Follow-up analyst notes from Enders Analysis or similar firms on Reddit's interest-based targeting resilience versus identity-based platforms
This Week in Reddit
👥 Reddit Tells Brands: Act Like a Citizen, Not a Billboard
At the Cannes Lions Festival, Reddit's executive VP of community, Laura Nestler, told marketers to "act like a citizen, not like a billboard" inside subreddits. Reddit's own guidance is blunt: brands that listen, contribute, and respect each community's culture consistently outperform those that treat Reddit like a standard social channel. The advice ties directly to how AI systems already read Reddit participation as a trust signal, not just a media buy.
🚫 Why Removing a Bad Reddit Thread Backfires
An SEO consultant's case study on a client whose branded search results were dominated by a thread titled "HORRIBLE": the fix wasn't a takedown request. Over six-plus months, the team built an official Reddit account, stood up a branded subreddit, and eventually got the original poster to try the product again and voluntarily pull the post. Mods treat removal requests as hostile, and the piece is a reminder that reputation on Reddit gets rebuilt, not deleted.
