Son zamanlarda üretken SEO’yu iten birçok yeni araç ve girişim görüyorum ve Chatgpt, Claude ve şaşırtıcı yüzey gibi LLM’lerin nasıl cevap verdiğini optimize ediyor. Saha, geleneksel SEO’nun ölmesidir ve bu görünürlük elde etmenin yeni yoludur.
Bu alanda yeni bir şirketin ortaya çıktığını görmediğim bir gün geçmiyor. Bazıları yasal geliyor, ama bir parçam harikalar… Bu sadece içerik pazarlamasının yeniden markası mı, yoksa gerçekten bir değişim var mı?
Başkalarının ne düşündüğünü merak ediyor ve buna dikkat ediyor musunuz? Yoksa hepsi yutturmaca mı?
It’s snake for noobs.
It’s just a new angle. Nothing else. Don’t be a victim of every random marketing copy or commercial
Edit: Some tools might help streamline your workflow or save time, but nothing thats absolutely necessary to get you top rankings or cited in LLMs
I don’t know if it’ll “completely change your SEO game” like some companies are saying, but I don’t think GSEO is something you can ignore anymore. In my opinion, it can’t hurt to optimize your content/website for it without going overboard. Generative AI searches for quality (or tries to), and it’ll only get better at doing that as time goes on.
The company I work for has started adding minor optimizations like key takeaways & summaries, and we focus more on content structure (which is just good SEO practice in general anyway) than we have in the past. I’ve read that it’s better to focus on quality answers & content than on stuffing keywords now. Is that true? Time will tell I guess.
To a certain extent, but it’s really no different than your normal SEO. If you search the same query twice in an AI that scrapes the web it will come up with different recommendations almost every prompt.
Its about understanding the “story telling” vs “Reality” and ky to reverse engineering or unlocking that is understanding the [query fan out](https://www.reddit.com/r/SEO/comments/1m1g8tp/community_llm_seo_discussion_the_query_fan_out/), which I posted here a few days ago and you need to read.
# Disinformation campaigns and conjecture
Wherever people are running disinformation campaigns without evidence – politics, cults, SEO, urban myths – people play to pseudo-logic and cognitive bias.
So – people try to paint a picture that “makes sense” – right now, the best picture is SEO is bad because it never recognized your content or branded strategy (why would it?) so LLMs are more intelligent and will reward you.
Most of its created by copywriters who lost work since LLMs… and here are the key red flags:
1. You need to write in a special way
Why? LLMs turn text into mathematical strings and optimize them via tokenization – why would you need to explain it? They are literally trained on platforms like X and reddit. They can lookup meanings of concepts in 0.000001 seconds.
Critical thinking: ask why do people go to an LLM to summarize or explain things but copywriters need to write ina sepcial way for LLms to understand and that LLMs seek out this content? Its ridiculously complex.
2. LLMs are not emulating google
LLMs cannot keep up with the compute power – processing, electrical usage, memory for their growing models. GooglePlex- which is just a static, 24-hour refreshed database is 1000X bigger,.
In other words LLMs cannot build a Google sized search DB AND build LLM compute power.
3. There just is no replacement for PageRank
content and brand marketers HATE PageRank because its objective – they think and want us to believe that LLMs are recognizing “great content” – I can’t tell you how naive and simplistic in a world where everything is subjective a viewpoint this is…
4. Query Fan outs explain brand visibility
In the link above – I share examples of how prompts become queries – and its lack of ranking in those queries = why your brand or site is invisble.
5. ask for examples, not conjecture
Gemini, ChatGPT and Perplexity are not build by the people trying to tell you how they work – a nice story that “makes sesne” isn’t the foundation for how these tools work.
Why can’t they give examples?
I have examples where traffic reports in Looker run to 500 pages of LLM traffic and showing the Query fan out and how the results exactly match between Google and Perplexity so taht you can do this yourself.
Evidence is everything
here’s an example from Perplexity: