Overview
The numbers? According to Similarweb, AI platforms generated an average of 770.7 million referral visits per month worldwide between June 2025 and May 2026. That’s +117.4% year over year.
Impressive. But what should we make of it, concretely? And how do you adapt your digital strategy without panicking?
1.AI referral traffic is exploding, but volume doesn't tell the whole story
At first glance, the story is simple. AI sends more visitors to websites, so you need to be part of it. Except that when you dig deeper, the picture becomes more nuanced.
In terms of volume, marketplaces come first with an average of 46.8 million AI referral visits per month. News and travel follow, at around 44.5 million each. Hardly surprising: searching for a product, getting informed, planning a trip, all of it naturally leads to an external site.
But the sectors that receive the most traffic aren't the ones growing the fastest. Beauty posts +312.5% year over year. Fashion, +278.2%. Marketplaces, already leading in volume, keep climbing anyway: +237.3%.
Why does this nuance matter? Because a percentage can impress without saying anything about actual traffic. A small sector can double in the blink of an eye from a tiny base. An established category, on the other hand, can gain millions of visits with more modest growth.
To put it simply: growth reveals momentum, volume reveals scale. On its own, neither is enough.
So, what should you do with these numbers? If you sell online, these visits may already be numerous enough to deserve regular monitoring. In another sector, the right question lies elsewhere: is this channel gaining weight with your audience, even if its contribution remains modest today?
And besides, receiving traffic is not the same as getting value from it. Someone who arrives after asking an AI for a detailed comparison may have a stronger intent than a visitor who is just browsing. Maybe. But that hypothesis has to be checked against real behavior and conversion data. Not against the referral source alone.
What to remember: before moving any budget, establish a baseline.
- What share of your traffic comes from AI platforms today?
- Which pages do these visitors land on?
- Do they browse, request information, convert?
Without these answers, even a channel in full explosion remains hard to judge.
2.Why some sectors take off faster than others
Imagine two questions asked to an AI assistant.
"Which news sources should I follow to understand what's happening in the world?"
"Which ingredients are best for sensitive skin?"
Both can lead to a visit. But the path between the question and the click isn't the same. In the first case, the person wants to read an article, verify a piece of information. In the second, they're looking for recommendations, comparing brands, refining a choice before buying.
That explains a good part of it. According to Similarweb, beauty is the fastest-growing category, followed by fashion. These worlds lend themselves perfectly to conversational discovery. People describe their needs, their preferences, their constraints, the result they hope for. No more searching for a generic product: they ask for a recommendation tailored to their situation.
A word of caution, though. AI assistants help structure the exploration, but they don't necessarily send the user to the brand's website. Depending on the query and the platform, the answer may cite a retailer, a publisher, a comparison site, or another source entirely.
Hence a more strategic question: when an AI answers a question about your market, which sources does it rely on? And are you one of them?
The answer may surprise you. A company can dominate its commercial keywords and remain invisible in AI-generated comparisons. Conversely, a niche site that is useful and well documented can join the conversation without showing the highest search volume.
That doesn't mean AI visibility replaces SEO. The two are linked. Technically accessible pages, useful content, a clear architecture, credible information: all of it still matters.
What changes is the perspective. It's no longer just about the queries that bring people to your site, but also about those that influence their decision before they even click.
A B2B example: does your site answer the real questions people ask when choosing a supplier? Implementation complexity, integration, costs, security, long-term support. On the beauty side: explanations about ingredients, product comparisons, specific customer needs.
The goal isn't to anticipate every question put to an AI. That would be unrealistic. It's to understand what your customers need to know, and to say it clearly enough to be useful.
A concrete starting point: reread your content through the eyes of someone who hasn't yet chosen to work with you.
- Does your site answer their questions?
- Are your claims backed up with proof?
- Is the difference from your competitors obvious?
These questions apply to traditional SEO too. What changes is what they make possible: connecting your expertise to the new ways of searching.
3.From AI visibility to results: what should you measure?
Being cited by an AI assistant feels like a victory. But what does that visibility actually produce?
A brand can appear in an answer without earning a single click. A user remembers its name, goes back to Google later, types the brand, visits the site. Another clicks the cited link right away. In your analytics tools, these two journeys have nothing in common. Yet everything started with the same interaction with the AI.
That's why measuring AI search requires looking beyond simply counting referral visits.
Two notions to distinguish:
- AI visibility: your brand appears in relevant answers, in the form of mentions or citations.
- AI referral traffic: the visits generated when someone clicks a link from an AI platform.
This may sound technical, but the consequences are very practical. If you only monitor referral sessions, you risk missing the AI recommendations that feed later visits through other channels. And conversely: more mentions doesn't prove more qualified leads, or more revenue.
Where to start?
- Look at the data you already have. Your analytics tools can identify visits coming from AI-associated domains, even if attribution remains imperfect. Examine landing pages, session quality, visitor actions.
- Test your visibility. Ask a selection of relevant questions on the platforms that matter to your audience. Does your brand appear? Which competitors are cited? Which sources? Are the answers accurate, and true to the expertise you want to show?
- Tie it all to your objectives. Qualified inquiries, product discovery, online sales, brand awareness: your indicators must reflect your objective, not a generic checklist.
A word of caution on step 2: these observations are indicators, not definitive rankings. An AI's answers vary depending on the platform, the wording of the question, the context. A single test isn't enough.
A relevant measurement framework could include:
- Visibility: brand mentions and citations across a consistent set of relevant queries.
- Traffic: AI referral sessions, landing pages, engagement.
- Business impact: inquiries, purchases or other conversions attributable to AI-driven visits, when the data allows.
- Competitive context: changes in the brands and sources appearing in the relevant generated answers.
None of these indicators tells the whole story. Together, however, they let you know whether AI search is becoming a real source of visibility, and whether your approach needs to evolve.
The same logic applies to investment. You don't necessarily need a separate AI strategy, or to overhaul your website because a report shows spectacular growth. What you need: to understand where you stand, spot the real gaps, and prioritize what serves your digital objectives.
The first step isn't to chase every new channel. It's to understand where your audience is looking for answers, and whether your brand helps them find the right ones.
Conclusion
The rise in AI referral traffic shows that the paths to online discovery are multiplying. Some sectors already receive massive volumes. Others are growing fast from lower levels. The real challenge for businesses? Reading these differences without confusing market trends with their own reality.
Traditional SEO remains a solid foundation. But it may no longer be enough to explain the entire journey through which a customer discovers a brand. Content quality, technical accessibility, expertise, measurement: all of that remains essential. What deserves more attention is how these elements feed your visibility across increasingly varied discovery experiences.
There is no magic formula for appearing in AI answers. And that is probably why a measured approach matters more than ever.
Rather than reacting to every new statistic, start with the essentials:
- Audit your content.
- Monitor the AI platforms that matter.
- Tie your visibility data to concrete results.
Over time, these foundations will tell you where to invest more.
Because the question is no longer just: how do I attract visitors to my site? It's also: how do I stay discoverable while internet users are still forming their questions and weighing their options?
Do you want to understand your brand's visibility in AI search?
At Eminence, we help businesses connect SEO, content, data and digital strategy to strengthen their online presence. Understanding how AI is transforming discovery is a good starting point for spotting new opportunities and deciding where to focus your efforts.
Contact Eminence to take stock of your digital visibility and identify your opportunities.
This is a translation of the restyled French version from my previous reply. If you'd rather have a translation of your original text instead, or a Word file, let me know.
Source
Similarweb — AI Referral Traffic by Industry: What 770 Million Monthly Visits Reveal
Contact us for more information.
These topics might interest you
