AI Is Changing Search: What The Data Shows
If search visibility drops, so does revenue for many public companies. There is now a clear link between organic visibility and stock prices. When Google tweaks its algorithms or launches new features powered by AI, some brands see a sharp fall in online visitors, and sometimes, the consequences go well past digital marketing. That is why it has never been more important to keep up with how AI is shifting search.
Many of the statistics and insights you will find online talk about which platforms send the most traffic, how often AI platforms cite certain sources, or what kinds of pages show up in those results. But if you actually run a website, or you work in SEO, what you need is clarity on where things stand now and what is likely to work as AI search grows.
So, here is a summary of key findings, based on my own research and industry data. We will go deeper in a moment, but the pattern is simple:
AI search tools, like ChatGPT, Google’s AI Overviews, and newer platforms, are now in the mix with traditional search engines. But they do not behave the same way, and the ecosystem is unpredictable for businesses and brands.
AI vs. Traditional Search: Where Traffic Really Comes From
Most referral visits still arrive from regular Google search. Even with the buzz around AI, traditional search engines drive most web traffic today. According to fresh industry estimates, Google still sends hundreds of times more visitors to sites than ChatGPT, Gemini (Google’s AI chat), and Perplexity combined.
If you watch your analytics, you probably see that too: organic Google visits make up a big chunk, while direct clicks from AI platforms are often just a trickle.
But ignore that trickle at your own risk. The share of visits from AI tools is growing fast. Over the last year, some sites have seen this segment increase by nine times, even if it is starting from a tiny base. In time, that trickle might become a steady stream.
There is another twist. Even though AI platforms send less traffic, those visits behave differently. In one study on SaaS signups, AI referrals made up only a small share of total visits, but those users converted to paying customers at vastly higher rates. Sometimes 20 times higher. What does it really mean? The quality of these visits might end up mattering even more than the total volume.
What Gets Cited in AI Search Results?
If you want to appear in an AI answer, or in Google’s AI Overviews, understanding what content gets cited is key.
The pattern that emerges is not always logical. For instance, the majority of sources cited in responses from platforms like ChatGPT, Perplexity, and Google’s AI Overviews do not match up perfectly with Google’s regular top ten rankings for the same terms. Somewhere around 80 percent of sources cited by these AI results are not in Google’s top ten at all.
You might expect this means organic SEO has lost influence, but that is not quite right. Pages that already get more organic search traffic are mentioned more in AI search, too. There is a connection, but it is not direct.
AI results favor sources that are fresh. In fact, AI answers, across platforms, tend to cite pages that are more recent than what shows up in regular search.
There is also a distinct shift in what sorts of sites show up. Google AI Overviews, for example, seem to pull a higher proportion of results from user-generated content sites, think Reddit, Quora, or YouTube, compared to their usual rankings. These platforms sometimes account for more than 5 percent of Overviews citations. ChatGPT, meanwhile, seems especially partial to news outlets like Reuters or Apple News.
Wikipedia is still a huge presence across all these tools, but ChatGPT is more likely to reference it (over 16 percent of the time), with Perplexity and AI Overviews lagging a bit behind.
Visibility: Why Some Brands Dominate in AI Search
One thing is clear, strong branding pays off. Brands that are broadly mentioned online, especially on highly linked pages, dominate visibility in AI Overviews. In some cases, the top 50 brands picked up almost a third of all AI Overview mentions.
But this result is not as obvious as you might think. While backlinks still help, the connection between classic link-building and AI Visibility is weaker than the link between simple online recognition (mentions). So, having your brand referenced, even without a link, counts.
It gets more complicated: over a quarter of brands have no representation in AI Overviews at all. Even if they rank well organically, they can be completely invisible in these new AI-driven features. Sometimes it is just because the brand has little coverage or is not seen as authoritative in enough places.
Branded web mentions have the strongest link to AI Overview visibility, followed by anchor text and search volume. Branded ad spend or traffic has a weaker connection.
Here is a quick snapshot:
| Factor | Correlation with AI Overview Mentions |
|---|---|
| Branded mentions | Strong |
| Branded anchors | Moderate |
| Backlinks | Weak |
| Branded ad spend | Very Weak |
The main point: no matter how well you rank, your brand must get referenced to show up often in AI summaries. If you are invisible online, these features might skip over you entirely.
Google AI Overviews: Reach and Impact
AI Overviews is probably the most visible AI-driven change in search right now. It shows up in about 16 percent of all US searches and has over a billion users each month. That is more than a quarter of internet users worldwide.
For publishers and website owners, the most noticeable effect is lower clicks. AI Overviews reduce organic clicks by about a third. Fewer people tap through to visit the actual sites.
Sometimes, even if your page ranks at the top, AI Overviews can appear above you, answering the user’s question right in the search results. In fact, almost 9 percent of Overviews show up outside of the top organic spot, sometimes as low as sixth position.
If you are trying to get cited in AI Overviews, the odds are best for pages with high organic rank already, about three-quarters of all Overview citations are pulled from Google’s existing top ten. But this is not always the case, and for some queries, Overviews pull in pages that do not rank well at all.
Some websites, like large user-driven communities or established media outlets, are disproportionately likely to be cited. For brands, making sure your content is trusted and referenced across other sites helps boost your odds.
One additional pattern: AI Overviews show up more often for longer queries and those with higher informational intent. If the query is short (or highly branded, or very local), you may not see an AI Overview at all.
ChatGPT, Perplexity, and Other AI Platforms: How They Choose Sources
ChatGPT leads the way in AI-powered referrals. Although the volume is a fraction of classic search, growth has been quick.
But there is a quirk, very little overlap with Google’s top results. Over 80 percent of sources cited by ChatGPT are not in the top ten for Google. Even stranger, some content cited does not appear in Google search at all.
ChatGPT cites newer content more aggressively, ordering references from the most recent to the oldest. News sources and timely pages feature more often.
YouTube and similar video-rich pages do better in other platforms, like Perplexity and Google Overviews, but not as much in ChatGPT.
If you work with health or medical content, it might be surprising to see that Mayo Clinic and Cleveland Clinic are cited often in Google’s AI Overviews, but rarely in other AI platforms.
Reddit fares best in Google Overviews. It rarely makes the top ten in ChatGPT or Perplexity.
Here is an interesting pattern that I think gets missed: ChatGPT is far more likely to feature news organizations (Reuters, for example), while Google’s AI Overviews tilt toward user-generated content. It might reflect their core data sources or how much they trust established brands.
How AI Bots Impact Websites
AI search platforms do not just attend to content, they crawl it actively with their own bots.
The number of AI bots crawling the web has doubled in the last year. There are at least 20 major ones now. The leader, GPTBot (the OpenAI crawler), sees the most active blocking from sites, almost 6 percent of all sites have blocked it outright.
Some publishers worry about server load and unauthorized content use. Blocking helps with that, but it may mean missing out on being cited in AI results.
AI bot crawling is now second only to regular search engine bots in terms of traffic. If you check your server logs, you might spot requests from names like GPTBot, ClaudeBot, or Google-Extended.
Does AI-Generated Content Help or Hurt SEO?
The bulk of new online articles and product descriptions now include at least some AI-generated text. About three-quarters of new web content (in several studies) is at least partly written with AI tools. That other quarter, written entirely by people, does not always perform better.
Is it a bad idea to use AI for content? It depends. Almost 87 percent of content in the top twenty Google results is at least partly AI-assisted.
But if you look for pure, 100 percent AI articles ranking at the very top spot, those are rare. Usually, the best rankings go to hybrid content, AI for structure, then written and edited by real experts.
There is one problem: human-only content has sometimes been more vulnerable to ranking drops in Google updates, by a small percentage, but it is worth considering. Perhaps because AI content follows templates that are less likely to stand out as “spammy” when algorithms change.
Who Is Using AI for Content, and How?
Marketers are using AI tools for nearly everything: writing blog posts, brainstorming ideas, outlining, and even updating old posts.
AI-generated blog posts make up the most common output. Marketers using AI publish around 40 percent more content each month. That is a big difference.
But hardly any companies just push out unreviewed AI content. Over 95 percent have some editing or review process; pure AI output is almost never published without a human scan.
The costs add up, but not as fast as hiring more writers. Average spend on AI tools for content creation is around $200 per month per company. That is less than the cost of even a single freelance post for many businesses.
Disclosure of AI use is rare. Fewer than one in five companies are open with readers that content was created or helped by AI.
As for which tools marketers use, ChatGPT remains the most popular, way ahead of competitors. Gemini and Claude are catching up but have not displaced it yet.
AI Traffic: Small Numbers, Big Results
The absolute number of visitors coming from AI platforms is still low. Web analytics suggest it is about 0.1 percent of total referral traffic.
But the effect can be outsized. In some SaaS scenarios, AI referrals made up only 0.5 percent of traffic but 12 percent of new signups. The conversion rate is much higher with these visitors, maybe because they arrive ready to take action.
Sessions from AI platforms are also different. Users view fewer pages per visit and bounce slightly more. Oddly, they tend to stay on the site for a few seconds longer. Maybe it is just people reading the answer before leaving.
Sites of all sizes receive AI traffic, but it is concentrated more on midsize and smaller sites, perhaps because the biggest, best-known brands have not been featured as heavily yet in these newer tools.
How to Prepare for More AI in Search
If you want your brand or site to be included in AI search results, what should you do differently?
Here are a few takeaways, based on what is working now:
- Focus on being mentioned online, not just linked. AI Overviews pick up branded mentions and references even without classic backlinks.
- Keep your content updated and relevant. AI platforms cite fresher content more often. Do not let your big posts age out.
- Build true authority. Wikipedia, major news outlets, trusted user forums, and reference sites are all frequently cited. Think about your content angle and fill knowledge gaps those sites have missed.
- Test AI-specific search. Try running typical queries in ChatGPT, Perplexity, and Google’s search and look for patterns. Are you cited? Who is, and why?
- Blend AI and human content. Pure AI-generated text rarely wins the top organic positions, but hybrid content does very well.
- Monitor your referral data for changes. If you notice more high-converting visitors with low overall visit counts, AI may be the source. Plan landing pages and CTAs to match.
Some tactics will feel more comfortable than others. I am a believer in testing, not following advice blindly. If something is not working for your niche, pivot.
Your Questions About AI and Search
Is it possible for a website to grow quickly just by publishing a lot of AI-written content?
In theory, yes. Some sites do see faster growth by scaling output with AI tools. But most of the time, the best results come from combining AI for structure and speed with expert review. If you publish nothing but raw AI output, the odds of sustained top rankings are lower. Quantity helps, but quality and authority still matter more.
What is the best way to get cited in Google’s AI Overviews?
There is no guaranteed method, but the strongest common factor is branded mentions across trusted external sites. Organic ranking helps, as does having fresh content. But, getting cited in Wikipedia, media outlets, or authoritative community sites seems to help the most.
Should every business invest in AI content tools now?
Maybe. If you care about scaling content or competing in search results where AI platforms matter, it is worth testing small investments first. Start with outlining, data research, or bulk drafts. Over time, adjust based on your real business outcomes.
Curious about how your site fares in AI search? Or if you are seeing referral spikes from new sources? Let me know what you are observing. Sometimes, the best insights appear in those weird periods between big industry updates.
Need a quick summary of this article? Choose your favorite AI tool below:
