Llms.txt duydunuz mu? AI modelleri için bir hile sayfası gibi. ChatGPT gibi araçları en önemli sayfalarınıza yönlendirmek için bu metin dosyasını sitenizin köküne yerleştirin. Bunu robots.txt’in genç kardeşi olarak düşünün. En iyi içeriğinizi engellemek yerine, en iyi içeriğinizi vurgular.
Yoast’ın neyin girdiğini seçmeme izin veren yeni manuel modunu test ediyorum.
So – instead of writing what your site is about – you have to create a text sitemap for a tool that can read and convert a websites content into a mathematical model in seconds – and somehoiw this “llms.txt” – which is neither a robots nor a sitemap despite the numerous spammy analogies – give you an edge over sites without one – because somehow spammers, scammers, low quality sites can’t jsut do it to get an edge too?
# The reason this exists?
The need to believe that PageRank SEO is being replaced by LLMs who appreciate people’s writing is greater than all the evidence in the world……. that people who do not understand circular logic: that a page or site which claims to be the best in class for any topic (because the owner believes it to be true) cannot be taken at face value because every other site is making the same claim?
JUst kidding – the answer clearl is a text file amid a whole site of HTML to halp a tool with 30 billion text references to stop from getting lost in a highly organised mathematical model of its own creation….
Makes total sense.
I don’t believe it’s actually being used by most AI crawlers. At least not yet. Would love to be proven wrong by as we’ve adopted it in our documentation software.
Actually I did try it for fun.
And then I had my team remove it when we rebuild the WP backend 3 weeks ago.
And we have continued gaining in ChatGPT – if not at a faster rate but it seems thats because we rank better in Google and ChatGPT is moving more and more to Google.
# Intentional vs Magici
Secondly, I dont believe in magic. And so I dont believe that Schema adds value to FAQs, articles, blogs – except things like flight, cinema, hotel data – which is hard to scrape but only useful for Google not LLMs.
And whats funny about optimism is that it sounds a lot like luck which sounds a lot like anti-science… I dont do SEO “hoping” our clients will rank or write blog posts “hoping” I’ll rank. I’m not saying I rank instantly – I’m saying its not accidental or co-incidental. Obviously there are lucky accidents – historically speaking.
Here’s my major **tl;dr part:**
# I’m not wondering how to rank in LLMs, I’m doing it
And so – all of my traffic from Perplexity, ChatGPT, Perplexity and Claude isn’t accidental.
I know that these are not research tools. I know that if peole ask an LLM “who’s an AI SEO expert or Top SEO Youtube channel or Top SEO Blog or SERP Report or King of SEO – I know that they will take the LLM at face value because they believe they are research units.
So I intentionally went out to rank for them and anytime I test them, 99% of the results — I am
So I dont care about LLMs.txt or schema or special “LLM tickle writing” …..
I’ve yet to hear from a credible SEO practitioner spending any time on this. Mostly because no LLMs actually use it.