Hasan Deniz

SEO Consultant & AIO & Automations || Crawlability & Indexability Expert

[Hasan Deniz @hasbehas]

Simply put..

[Tim Soulo 🇺🇦 @timsoulo]

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.

[Sir Escanor (𝘏𝘰𝘱𝘪𝘶𝘮 𝘚𝘭𝘢𝘺𝘦𝘳) @EscanorReloaded]

CEOs are quietly realizing the AI replacement plan has a problem.
Two problems, actually.

One: the token costs for running AI agents are now exceeding what they were paying the employees they fired.

Two: when the tokens run out, the AI stops. Just stops. No continuity. No workaround. Just a spinning wheel where your workforce used to be.

You fired humans to save money and bought a subscription that bills you into a corner.

The employees you let go knew what to do when things broke.
The AI just invoices you for the outage.

And then there’s the permission problem nobody wants to talk about.

To do its job, the AI agent needs access. Full access. Your systems, your patents, your contracts, your future plans. Everything you spent years building, handed over to a process that has no loyalty, no discretion, and no skin in the game.

You didn’t hire a replacement.
You gave a stranger with no soul the keys to everything you own.

Enjoy.

[Hasan Deniz @hasbehas]

We are indeed living in the future... 😌

https://t.co/wvuVDjN9oU

[Ryan Jones @RyanJones]

companies trying to replace headcount with AI instead of enabling employees with AI are going to regret it. @dohertyjf #SEOweek

[Hasan Deniz @hasbehas]

@RyanJones @ChrisLaubAI @lightpanda_io well, I did test obscura and Lightpanda is several times better in multiple ways...

[Hasan Deniz @hasbehas]

@addyosmani If you are investing in your own AI setup and it feels invisible, this is for you. 👇

Keep going! One day your own tool tells you where you stand, and you realise the quiet work was the work!

[Hasan Deniz @hasbehas]

Off-the-shelf AI skills will always lose to the ones you build around your own work. Here is how I found out!

I asked my Claude to go through Addy Osmani's agent-skills repo (in 1st comment) and compare it against what I already have. +

[Hasan Deniz @hasbehas]

@addyosmani Because on that one narrow thing, my 2+ years of small moves had quietly stacked higher than the public teaching level. Rules, memory files, voice guides, KB routes, banned words, project-level overrides. Nothing shiny on any given day!
+

[Ryan Jones @RyanJones]

This is kind of scary. I guess great for advertisers though.

[Vaishnavi @_vmlops]

GOOGLE JUST GAVE AI AGENTS THE FULL POWER OF CHROME DEVTOOLS

your ai coding agent can now open a real chrome browser, click around, inspect network requests, take screenshots, record performance traces, run lighthouse audits, and read console errors all through mcp

debugging a slow page? it records a trace and gives you actionable insights.

weird network request? it lists them all with full details.

console errors with garbled stack traces? source-mapped and readable.

one `npx` command. works with cursor, vs code, windsurf, gemini cli, and more

this is what browser debugging looks like when your ai agent has devtools access

https://t.co/vol59RnMUW

[Hasan Deniz @hasbehas]

ouch....

[Ryan Jones @RyanJones]

I can't wait until it all comes crashing down on the fully automated AI SEOs. I don't care if "but it's working"

At the end of the day, no user wants AI generated content from your website. If they want AI generated stuff they'll go directly to the AI.

Eventually the search engines will figure it out and these sites will all fail. It's not sustainable.

[Aleyda Solis 🕊️ @aleyda]

Sad for the businesses that don’t know better and fall for this.

Modern search engines like Google value originality, experience, and expertise in content that is human first. That’s how they ensure the best user experience.

Generating content at scale without these is a recipe for short lived success and a mess to clean up later to recover rankings.

[Marie Haynes @Marie_Haynes]

Woh. Search console has been inaccurately reporting impressions since May 2025.

A fix is coming over the next few weeks.

https://t.co/EJyx8tVh3q

[G 💙🧡 @iamgstv]

@sengineland We are stuck in this AI, AEO, GEO nonsense. People forget core fundamentals like a proper hierarchy of your sitemap. Or the ability to crawl your site properly.

[Hasan Deniz @hasbehas]

This is gold!
I conducted several tests, and unfortunately, not all LLMs/agents are able to read the page as intended.

[Jan-Willem Bobbink @jbobbink]

I tested what AI agents can actually read on e-commerce sites. The gap between what humans see and what agents see is bigger than most teams realize.

Everyone is talking about GEO and AI search optimization. But most of the conversation focuses on content strategy and citation patterns. Almost nobody is talking about the technical layer underneath. And that layer is where most sites are silently failing.

When a human visits a product page, everything works. You see pricing, stock status, reviews, size guides, shipping info. You select a variant and the price updates. The experience feels complete because your browser runs JavaScript and renders everything on the fly.

AI agents do not get that experience. Most of them cannot execute JavaScript. They read the raw HTML before any client-side rendering happens. And on a growing number of sites, that raw HTML is half-empty.

Here is what typically breaks. Search functionality built entirely in JS, so agents cannot discover products the way users do. Product variant selectors where size, color, or flavor options only load after a user interaction, so agents never see the full range or pricing per variant. Review widgets from third parties like Trustpilot or Bazaarvoice that inject ratings client-side. FAQ accordions where answers are hidden until a user interacts. Faceted navigation that filters without changing the URL. Shipping policies rendered inside JS modals. Structured data generated by JavaScript instead of served in the HTML source.

Product variations deserve special attention. Many e-commerce platforms handle variants entirely client-side. The default HTML might show a single SKU with a base price, but the full catalog of options and their pricing only appear once a user makes a selection. For an AI agent trying to recommend a product in a specific size or configuration, that information does not exist.

The result is that AI agents see a stripped-down version of your site. They miss pricing, reviews, specs, and your full product range. All the information that makes your page useful to a human is invisible to the systems increasingly deciding which brands get recommended.

The brands that win in AI-driven discovery will not just have the best content. They will be the ones whose content is actually accessible when a machine reads the page. Server-side rendering, clean HTML fallbacks, and structured data in the source are the foundation of being visible in an AI-first world.

If your product information requires JS execution to appear, it does not exist for all AI agents.

[Mo @atmoio]

AI is making CEOs delusional