AI Traffic Is Up 1,324%: What Semrush's 126M-Prompt Study Reveals About Your Brand's Visibility Gap

Here's something that should bother you: AI-referred traffic to retail websites grew by 1,324% between October 2024 and May 2026. For travel sites, the number is 2,215%. And yet 45% of marketing leaders say they still can't accurately measure their brand's visibility inside AI-generated answers. That's not a minor gap in analytics tooling — that's a massive, fast-growing channel flying completely blind for almost half the industry.

Semrush (now under Adobe ownership after a $1.9 billion acquisition) just dropped the expanded 2026 AI Visibility Index, built on analysis of 126 million U.S. AI search prompts collected between January and April 2026 across ChatGPT, Google Gemini, Google AI Mode, and Google AI Overviews. I've been going through it, and there are a few findings in here that I think most SEO teams are going to find uncomfortable — in a productive way.

The 81% vs. 36% Number Everyone Should Pin to Their Wall

The headline stat from this research: 81% of organizations that integrate SEO and AI visibility into a single workflow reported increased traffic or leads from AI platforms. Among teams that manage the two separately — your traditional SEO team over here, some newer "AI search" task force over there — only 36% reported the same result.

That's a 45-point gap. And it doesn't come from having better technology or a bigger budget. It comes from organizational structure. How your team is set up. Whether the person optimizing your web pages is in the same room — literal or metaphorical — as the person thinking about why ChatGPT keeps recommending your competitor instead of you.

I get why companies siloed these things initially. AI search felt like a different beast. You had your Google rankings in one dashboard and some vague question of "how do we show up in ChatGPT" in another conversation entirely. But this data pretty clearly shows that treating them as separate disciplines is costing you. The playbook isn't entirely different — it's an extension of the same fundamentals — and the teams who figured that out early are already pulling ahead.

Being Mentioned Is Not the Same as Being Cited

This is probably the most important conceptual shift the report surfaces, and it's one I haven't seen articulated this cleanly before: there's a hard difference between an AI mentioning your brand in a response and an AI actually citing your website as a source.

On Gemini specifically, the overlap between brands that get mentioned and the actual domains that get cited can be as low as 30%. Which means for every ten times Gemini talks about your brand, it's pulling the actual supporting content from somewhere else — a review site, a Reddit thread, an industry publication — seven times. You're in the conversation, but you're not the authority in the room.

Why This Matters Practically

If you're only tracking whether your brand name appears in AI answers, you're measuring the wrong thing. Getting mentioned but not cited means you're dependent on third parties to accurately represent you — and you have no control over what they say or how up-to-date it is. The brands consistently winning here have both: they're getting mentioned and they're earning the citation. That requires content structured to be citable — specific, factual, sourced, and authoritative — not just content that talks about what your brand does.

ChatGPT vs. Gemini: Two Very Different Citation Behaviors

One thing this index makes clear is that "AI search" isn't a monolithic thing. The four platforms in the study behave very differently, and what works on one won't necessarily translate to the others.

ChatGPT cites an average of 15 sources per response. It casts a wide net, pulling heavily from community and reference platforms — Reddit and Wikipedia are cited constantly. Gemini, by contrast, averages just 3 citations per response. It's extremely selective.

The practical implication: a brand whose visibility strategy relies on breadth — lots of mentions across lots of forums and publications — will naturally perform better in ChatGPT's wider net than in Gemini's narrower one. If you're only measuring "AI visibility" in the aggregate, you might be getting a false positive on one platform while being nearly invisible on the other.

The "Universal 36" Problem

Out of more than 1,200 brands Semrush tracked across all four platforms throughout the study window, only 36 maintained top-100 visibility on every platform, every single month. Semrush calls them the "Universal 36" — YouTube, Amazon, Walmart, that tier.

The uncomfortable reality is that for everyone who isn't already a household name with massive cross-web presence, platform-level differences matter enormously. You can't just pick one AI platform to "optimize for" and call it done. You need to understand your citation profile on each one separately.

Where Your AI Narrative Is Actually Coming From

Another finding that's worth sitting with: your AI narrative isn't primarily shaped by your own website anymore. AI platforms increasingly pull from customer reviews, community discussions, independent publishers, retailers, and vertical-specific sources to understand and describe your brand.

The Patagonia case study in the index is instructive here. Patagonia maintained an AI visibility score in the high 70s to low 80s throughout the study — and Semrush attributes this less to their website content than to a network of third-party outdoor gear review sites (OutdoorGearLab, REI, Switchback Travel, GearJunkie) and sustained Reddit presence. Their owned content didn't get them there. Their earned presence did.

This is the part that traditional SEO teams often don't own. Link building focuses on authority metrics, not on the quality of narrative those links carry into AI training data and retrieval systems. PR and comms teams have historically focused on reach and coverage count, not on whether the coverage is structured in a way that AI can actually extract and trust. That gap is the gap the 81% vs. 36% split is really describing.

Industry Concentration: Where the Opportunity Actually Lives

The index breaks out competitive concentration by vertical, and there's some genuinely useful signal here for deciding where to double down.

In News and Media, the top three most visible brands account for 82.9% of total category visibility. Consumer Electronics is similar: 76.9% for the top three. These categories are locked up. If you're a mid-size player trying to build AI visibility in those verticals, you're fighting a very uphill battle against entrenched dominance.

Finance and Industrial tell a completely different story. The top three brands in Finance account for just 41.4% of category visibility — Industrial is 42.2%. These are much more fragmented categories where consistent, credible content can actually move the needle for brands that aren't household names yet. If you're in a technical, B2B, or finance-adjacent space, the AI visibility opportunity is more open than you might think.

The Measurement Problem Isn't Going Away on Its Own

That 45% figure — nearly half of marketing leaders unable to accurately measure their AI visibility — is partly a tooling problem and partly a prioritization problem. Most GA4 setups don't differentiate AI-referred traffic clearly. Most SEO dashboards don't track citation patterns across ChatGPT, Gemini, and AI Overviews simultaneously. And most attribution models weren't built with AI-mediated discovery in mind.

Separate Similarweb research found that AI-recommended brands were 2.5x more likely to receive a website visit within seven days of a recommendation — but standard analytics capture almost none of the interaction chain that led to that visit. So you're getting the traffic, but you don't know what triggered it, which means you can't replicate it.

The tools are getting better. Semrush's own AI visibility toolkit now tracks across all the major platforms. Google Search Console added AI Impressions reporting earlier this year. But none of this helps if your team's structure means the data lives in a separate silo from whoever's making content decisions. Which brings us back to that 81% vs. 36% split.

Bottom Line

The Semrush AI Visibility Index isn't just another "AI is changing SEO" report — it's got actual numbers behind it, and those numbers point to a pretty specific problem. If your SEO team and your "AI visibility" efforts are two separate conversations, you're losing ground. If you're tracking brand mentions without tracking citations, you're measuring the wrong thing. And if you're treating ChatGPT performance as a proxy for Gemini performance, you're flying blind on one of the three most important platforms in AI search right now.

Practically, here's where to start: audit your citation profile on each AI platform separately, identify which third-party sources AI is actually pulling from when it discusses your brand, and get your content structured with the kind of specific, quotable, factual depth that makes it citable — not just findable. The teams doing all of this together, in one coordinated workflow, are the ones showing up in that 81%.