Google claims that AI in search is not stealing traffic, it drives more queries
Google has published an article that says AI is not stealing traffic from websites. In fact, it is driving more queries and higher quality clicks. We believe you!
The Mountain View company has highlighted how AI Mode, AI Overviews have improved the user experience, and that people are searching more than ever, by asking more complex questions.
Liz Reid VP, Head of Google Search, writes at the Google blog, saying that according to the company's data, that search volume to websites has been relatively stable. She also says that the average click quality has increased, i.e. users click on results instead of clicking back.
These claims are in stark contrast with studies published by third-parties, a fact that has been acknowledged by Google. Pew Research Center, which I mentioned in a previous article, said that Google users who read a summary generated by an AI, are less likely to click on links to websites. They mostly just close the tab, or the browser, once they have the information they wanted. This might benefit search engines, but has a serious impact on blogs, which depend on web traffic for earning revenue.
Google has denied that AI Overviews has resulted in a dramatic decline in traffic, and says that third-party reports are based on flawed methodologies, isolated examples, or even based on data before the roll-out of AI features in search.
"Our data shows people are happier with the experience more", "total organic click volume from Google Search to websites has been relatively stable", "average click quality has increased". Those are pretty words. But that's the problem, as TechCrunch notes, we just have to take Google's word for these. You talk about incorrect reports and flawed data, but don't provide your own data to back your statements.
I liked this part of Google's article which reads "While overall traffic to sites is relatively stable, the web is vast, and user trends are shifting traffic to different sites, resulting in decreased traffic to some sites and increased traffic to others." Well, if that's not a hint about AI's impact on search, I'm not sure what it is.
It's not just Google though, most AI services are affecting blogs and other websites negatively. Very few people actually bother to click on the links that are quoted by chatbots. Well, you should, not as a courtesy to the website that helped you, but at least to verify that the information provided by the chatbot was actually taken from the web page. Sometimes, AI hallucinates with incorrect links.
AI isn't sentient, it just regurgitates what it finds on the internet. People find it convenient to use, they don't realize the impact it has on their privacy, and willfully provide data that is used to train large language models. This data is used by companies to make money. Blogs are also being exploited as a resource in this process.
My question is this. Without the sources, without the websites that feed the AI the information that it provides, Can AI exist? Everyone is focused on short-term profits, nobody thinks about the long-term impact.
Google is facing an antitrust lawsuit against publishers in Europe for anticompetitive practices, and is under scrutiny in the U.S. in a lawsuit against the Department of Justice.
Do you use AI for search? Do you follow up by clicking on the links?
Whenever Google/Alphabet claims they “have improved the user experience” the exact opposite is true based on all the user-commentary you can find everywhere…
Google’s new tag-line should be “F*ck the user – we only want their personal data”
“Google users who read a summary generated by an AI, are less likely to click on links to websites. They mostly just close the tab, or the browser, once they have the information they wanted. This might benefit search engines, but has a serious impact on blogs, which depend on web traffic for earning revenue.”
Yes, it’s the new norm. I use DDG Assist all the time, and I find it far superior to clicking through websites trying to find a simple answer. When things get technical, I go to ChatGPT.
“Changing search patterns have brought about “decreased traffic to some sites and increased traffic to others,” with websites with forums, videos, podcasts, and “authentic voices” benefitting the most, according to Reid. Users are also more likely to visit sites containing in-depth reviews, original posts, or thoughtful analyses.”
Sort of a Spencerian “survival of the best, the brightest, and the most talented.” A “real” site with “real” writers who produce “real” content will flourish.
“Well, you should, not as a courtesy to the website that helped you, but at least to verify that the information provided by the chatbot was actually taken from the web page. Sometimes, AI hallucinates with incorrect links.”
Don’t even bother wasting my time! Why should I entertain the idea that the information provided on a website is correct? In all honestly, I’m not into “fact checking” AI responses. If I do a fact check, I don’t go to the provided website. I find a completely different site to verify the information.
AI really shines in tech answers!
@VioletMoon,
> “AI really shines in tech answers!”
You won’t be surprised that I consider this enthusiastic assertion as arguable if you remember my positions regarding AI brought to the masses and in particular within search engines.
I believe quick answers are profitable for no one in proportion of the question’s implications in terms of elaboration and complexity.
– If questions are simple, basic and in particular related to facts (but not their origins) then a search engine will provide most of the time in its first results links to pages which answer that question… and more, or answers to linked topics.
– If questions are complex search engines will summarize that complexity. I don’t think complexity may be summarized.
– The number of things we discover and learn when searching for answers and landing on related topics is fascinating, in the same way we’d discover old notebooks and photos when searching in the cellar or basement for that old magazine we’re sure to have kept somewhere. See what I mean? Reducing the benefits of searches’ side-adventures by being fed by an AI summary of strict answers is not the way to go IMO. I could write a whole chapter with references, but this is not the place.
– To end with an agreement : AI for the masses, that is for each of us (companies and science is another area), may be of great help when it comes to organizing one’s data, elaborating spread-sheets for a given scheme … but, still : that was the argument of those who worshiped the first home calculators… and now, ask kids to calculate with a pen and paper, not to mention mentally: result is terrifying. Ask ’em kids to send/receive emails, to manage folders on a computer, many just don’t know because what they want are places (social sites) to debate of things they ignore and are likely to never know in-depth if they never read (I mean books) and get information either from these social environments either from AI which will answer quickly as they like it (“no fuss, no point in loosing time digging when AI has it in a few words), hence do an arguable work but, moreover, a work in place of the user. And that includes home-work and even academic thesis.
AI does not shine, I’d rather say it’s raining AI like never before.
Interesting perspective from Google, but as someone who works with businesses on their digital presence, I’ve seen firsthand that AI-driven search can be a double-edged sword. While users may enjoy faster answers, many small and mid-sized companies depend heavily on organic clicks to reach potential customers. If AI Overviews reduce the need to visit actual websites, that could slowly erode the value of producing in-depth, quality content.
IMO journalists should to stop using the fluffy terminology espoused by AI companies such as “hallucinations” and call it what it really is: a common tendency to fabricate misinformation.
Don’t allow AI companies to steer the narrative.
I rarely defend corporations, but I do not see much difference between “hallucination” and fabricated misinformation.
I remember watching a movie and falling asleep during it. In my infinite wisdom, I decided to write the review after waking up. It turns out that I wrote parts of my dream into review. Of course, I was taken to task for fabricating misinformation which came from my sleep hallucination. So they kind of the same thing or one causing another.
Perhaps English isn’t your first language. In which case, maybe it would be better to make it abundantly clear and replace “hallucinations” with “false information the AI model completely made up”.
@ECJ, be noted that “false information the AI model completely made up” is a characteristic of hallucinations provided it was not intentional, deliberate. Maybe did you mean that AI is programmed to do so?
Believe me, I’m no fan of AI but as much as I can understand the pertinence of disinformation and intoxication, as little can I perceive the sens, the advantage of of distributing false information characterized as an hallucination.
Why is the word ‘hallucination’ evoked? Simply because information provided defies the common sens. When you want a lie to be efficient you try to make it plausible, it has to be understood rationally and/or emotionally, you won’t cry it out insanely, right? This is why AI hallucinations do exist, it is for the benefit of no one, it is only an issue.
Generally speaking and as always we assist to two approaches, one which worships AI and another which blends it with conspiracy. Perhaps as always a middle approach is a better one, otherwise we, human beings, may very well, ourselves, hallucinate :)
I’ve stopped using Google because it’s so heavily censored. Yandex is more like the old Google. I can actually find what I’m looking for.
It’s only good for some things, but not for all. Some different searches require different search engine. The idea that one search engine can serve you for all your searches is unrealistic.
Yandex is good to search for “high seas” things, but if you want to search for technical issues or more mainstream things, it won’t be so helpful.
“most AI services are affecting blogs and other websites negatively.”
Affecting indeed in that “They [Google users] mostly just close the tab, or the browser, once they have the information they wanted.” but also, as stated by the developer of Anubis [https://xeiaso.net/blog/2025/anubis/]:
“AI scrapers have been bullying the internet into oblivion and there’s not much we can do about it. The well-behaved bots will relent when you ask them to, add entries to your robots.txt (even though they should understand the intent behind a wildcard), or block their user agents.
A majority of the AI scrapers are not well-behaved, and they will ignore your robots.txt, ignore your User-Agent blocks, and ignore your X-Robots-Tag headers. They will scrape your site until it falls over, and then they will scrape it some more. They will click every link on every link on every link viewing the same pages over and over and over and over. Some of them will even click on the same link multiple times in the same second. It’s madness and unsustainable.”
When Google starts explaining that what they consider as a rumor, or a myth, is unfounded, it means they are aware of a true problem and start being on the defensive, which is seldom concerning that company (as other major ones), not to be mistaken about a company reporting on technical issues. Google is very “Don’t explain don’t complain” as long as cash is not in danger; otherwise, they deny, play violin and as a last resort: bite.
Not the first time Google has destroyed websites. First it was their algorithms. Now it’s their AI. Google has always claimed innocence, helpfulness, and good intentions. Remember, Do No Harm.