"Ranking #1 in Google is not enough for AI search"
...6 months later..
"The most strongly correlated factor with appearing in AI answers is ranking #1 on Google" lol
In the last 6 months at @Ahrefs, we analyzed over 1 billion data points across 14 studies. Here's what we learned about AI search optimization:
1) "Best X" blog listicles are the single most prominent content format cited by AI chatbots. They make up 43.8% of all page types cited by ChatGPT specifically.
2) 67% of ChatGPT's top 1,000 citations come from sources marketers can't influence: Wikipedia (29.7%), homepages (23.8%), app stores (6.6%). Only 32.3% are influenceable content like educational pages, reviews, news, and blog posts.
3) 28.3% of ChatGPT's most-cited pages have zero Google organic visibility. These pages get cited repeatedly by ChatGPT despite not ranking in Google at all. A completely separate discovery layer.
4) ChatGPT only cites about 50% of the URLs it retrieves. It fetches dozens of pages per query but uses half as background context without attribution. This means that being retrieved and being cited are very different things.
5) Adding schema markup had zero meaningful impact on AI citations. AI Overviews actually dipped −4.6%, while AI Mode (+2.4%) and ChatGPT (+2.2%) showed changes indistinguishable from zero.
6) YouTube mentions have the highest correlation (0.737) with AI brand visibility out of all the factors we studied (including all the conventional SEO metrics like backlinks, page count, DR, etc). This held true for both Google-owned and OpenAI products.
7) AI Overviews reduce clicks to the #1 result by 58%. That’s up from 34.5% just 10 months earlier. The trend is accelerating.
8) 99.9% of AI Overviews appear on informational intent queries. Transactional, navigational, and local searches are almost entirely AIO-free. Shopping triggers AIOs just 3.2% of the time.
9) For a given search query, Google’s AI Mode and AI Overviews reach the same conclusions 86% of the time — but cite almost entirely different sources (only 13.7% citation overlap).
10) AI Overviews change every 2.15 days on average, with 70% of content differing between consecutive observations. But semantic similarity stays at 0.95. The words, sources, and entities constantly shuffle, but the actual meaning barely moves.
The biggest AI misconception is that it saves time automatically.
It doesn't.
AI gives you answers faster.
You still need enough experience to know whether the answer is good, bad, or dangerously convincing.
@MozharAlexandr@prashant_baldha@WordPress Impossible to really know, but my guess is AI bypassing websites that don't need a CMS, in my 15 years building and supporting micro business websites virtually none actually need a CMS, we actually sold WordPress on the basis that it is so popular you are not tied to a developer
there’s no way this is real oh my fucking god
tldr: AI support system accepts AI generated video of the users profile picture for email change and password reset. Great work as usual Meta
My long covid started with psychosis around the six week mark of my infection. I was in the hospital for 14 days. So whenever I see a paper like this where they show generation of anti-NMDAR auto-antibodies, which are associated with anti-NMDAR autoimmune encephalitis, it always makes me wonder if that's what happened to me.
When I came out of the hospital they put me under the care of a psychiatrist. Nobody thought it had anything to do with long covid, they just treated me like a mental health patient. I felt awful, I was having muscle facilitations, sleep difficulties, difficulty with hand/eye coordination. I went and had a few private MRIs, which the psychiatrist took as a sign of me "regressing", that is she said I was delusional for trying to figure out what was wrong with me since the doctors already determined it (they didn't, they defaulted to some arbitrary 'depression', while also admitting they couldn't find the reason). And of course she tried to increase my meds of anti-psychotics so I'd stop trying to figure out my own health condition.
That's one area of medicine I think someone should take a wrecking ball to. The idea that someone can diagnose you with a mental illness with zero blood work, zero diagnostic markers, just one person's opinion, which can be absolutely life changing, is ludicrous. Even now I have the opinions of multiple experts including neuro-immunologists that that event was most likely related to the covid19 infection, it's still hard to change the official narrative in my file.
@jdevalk I'd love to data on llms.txt being actively hit... every study I've seen shows it doesn't. We'll have to agree to disagree. Well done on producing a genuinely useful site, I do appreciate your work
@jdevalk The listicle bit was sarcasm(/s) but a listicle can help you more than llms.txt. My point is that you have it down as recommended on your checklist when it does nothing for the client
@jdevalk Thanks for the article it was an interesting read. I would counter with, how would you answer a client question like: why are you spending your time on llms.txt instead of something that helps my business today.
The seo could have created another listicle instead of llms.txt /s
Milestone: 10,000 subscribers on my YouTube channel AI Coding Daily!
Took "just" 155 videos and 10 months.
But I think this is just the beginning.
Subscribe if you haven't yet!
https://t.co/QJb3TTNeAA
@rossstevens_uk I don't feel like I'm getting outside of my echo chamber when I dip into the main feed, it just seems like a cesspit of garbage, even the posts from "my side" are garbage.
@DuaneStorey Another interesting note is that the CEO runs the research group in Exeter that published the paper and most of the other authors have some sort of financial relationship with the company. Not that this is unusual... just interesting