Matteo Giannone

Entrepreneur, SEO expert, iGaming & online marketing enthusiast.
Having fun with technical SEO & {good,bad} link analysis.

[Matteo Giannone @matteogiannone]

@glenngabe The situation is even worse in some situations where with a single takedown MANY pages are removed.
In YMYL niches DMCA + fake trustpilot reviews are a real pain since many many months.

[Matteo Giannone @matteogiannone]

@LinuxHandbook Linux From Scratch (LFS) is a must if you want to understand which will be your favourite distro.
Mine has become Gentoo

[Matteo Giannone @matteogiannone]

@googlesearchc Hope just for US, Italy is heavily polluted with spam for YMYL queries

[Matteo Giannone @matteogiannone]

No AEO/GEO without SEO

[Harpreet @harpreetchatha_]

25 AEO, GEO, LLM SEO facts 📠 (to stop you from wasting money on something that doesn’t work)

1. AI visibility can refer to mentions or citations. (Mentions = being named by AI in the answer, citations = being linked).

2. SEO directly influences AI citations. Want to show up for “explain x”. Well you need content that explains x so your website can be cited. Good luck being cited for anything competitive without implementing SEO basics

3. SEO, marketing, your reputation & just doing business can ALL influence AI mentions. Want to be known for something? Well you need to say it on your owned assets and then other sites need to say it too. think of AI as a parrot 🦜

4. the latter 3 can also influence SEO, but that’s for another day.

5. Being mentioned does not = automatic success. The clickable links within AI may lead websites that don’t link to you. This increases friction for customers to visit your website. (Also, if websites don’t link to you and just mention you, how will you be found) [✨links are still important ✨]

6. Being cited does not = automatic ai success. You could be the 17th citation (that’s like being on page 6 of Google). Doesn’t mean anything.

7. AI mentions and citations are infinite. You can be mentioned and cited for an infinite number of queries. Likewise, you can track an infinite number of prompts 🌌 (track your main keyword within a prompt, you’ll be fine)

8. One highly successful TOFU webpage could have 100s of citations in an AEO tool (what is y, why explained, explain y so a 5 year old can understand).

9. One highly successful BOFU page may only have a few citations (y software, what’s the best y, what’s the best y product that solves for z)

10. No one on this earth has proven ROI after implementing an llms.txt file

11. No one on this earth has proven ROI after implementing schema for LLMs

12. Being on Reddit does not guarantee AI visibility. Being mentioned on Reddit in a positive or negative way can influence sentiment of your brand in AI answers

13. Different AI systems behave differently. Showing up in Google AI Overviews does not automatically mean visibility in ChatGPT, Claude, Perplexity or Gemini

14. Don’t sacrifice your Google rankings just to chase AI visibility. It could take 2, 3 years to recover

15. AI search with web search enabled uses data from search engines to form an answer

16. It can take years to influence true LLM training data

17. A strong brand can outperform a technically “better optimized” website in AI answers because LLMs take into account the stuff mentioned in point 3

18. You cannot optimize for every prompt variation individually. Focus on topical depth and building out the stuff in point 3

19. Remember AI being a parrot? 🦜 AI systems can confidently repeat incorrect or outdated information if enough sources on the web repeat it, so control the narrative

20. You don’t need to chunk your content. Just make your content readable

21. Every heading does not need to be a question

22. there is no universally accepted attribution model for AI search. A user may discover a brand in AI, then convert weeks later through branded search, direct traffic or another channel.

23. That said, AI Search is both a brand and performance marketing channel

24. You don’t need a prompt tracking solution if you’re just getting started. Ask AI questions about your brand - now you have something to work with.

25. Mass publishing content is not the answer to winning in ai search

What did I miss?

[Matteo Giannone @matteogiannone]

@zeeshana07x @glenngabe Don't cheat, just work.

[Matteo Giannone @matteogiannone]

@lilyraynyc @glenngabe @rustybrick do you see anything similar in other countries? Never seen this level of manipulation in my 20 years in the SEO industry, it's ridiculous and fascinating at the same time.

[Matteo Giannone @matteogiannone]

Google YMYL SERPs being aggressively manipulated by spam sites: exact-match domains, odd TLDs, expired sites, and obvious affiliate pages ranking where useful results should be. ChatGPT classified 6 of the top 10 results as spam in 10s. Please rebrand to Google AI Stupidity Mode.

[Matteo Giannone @matteogiannone]

ChatGPT can: why can’t Google? It feels like Google is no longer seriously interested in applying effective anti-spam filters. And if this is the future of “AI-powered search,” maybe we should stop calling it artificial intelligence and start calling it artificial stupidity.

[Matteo Giannone @matteogiannone]

Google YMYL SERPs being aggressively manipulated by spam sites: exact-match domains, odd TLDs, expired sites, and obvious affiliate pages ranking where useful results should be. ChatGPT classified 6 of the top 10 results as spam in 10s. Please rebrand to Google AI Stupidity Mode.

[Matteo Giannone @matteogiannone]

@glenngabe In Italy we're seeing more and more hacked sites and expired domains in top10 for the most competitive YMYL queries. Parasites are still dominating. Expired domains are pushed by high authorities links (expired domains as well) with exact kw anchor. Back to 2000s

[Matteo Giannone @matteogiannone]

After March 2026 spam update, YMYL SERPs are dominated by hacked sites, parasite and site reputation abuse sites. AI world is a nightmare, @Google forgot keeping human-in-the-loop , please take human quality raters back! Good job @googlesearchc #google #seo #parasite #GEO

[Matteo Giannone @matteogiannone]

@googlesearchc This spam update seems amazing, never seen an update rewarding so many spam and hacked sites so fast!!! 🍾

[Matteo Giannone @matteogiannone]

@fungibletokn @rustybrick It's common during core updates, seems like Googlebot relies on an old (and outdated) dataset first, then it moves to more recent data. Nothing new.

[Lily Ray 😏 @lilyraynyc]

Let me be very clear about the Helpful Content Update.

When Google announced this update, THEY were the ones who went on record claiming that recovery is possible, and that sites could begin to regain traffic after making changes and improvements that would lift the classifier that suppressed sites’ traffic.

In fact, there is even a recording of Google originally stating that recovery was possible “within in a few weeks.”

Of course, that language gradually changed to “a few months” and eventually, even more ambiguous timelines, leading up to now - where Google has said at multiple Creator Summits - including one a couple weeks ago - that “we are still working on this and are taking it seriously; it’s not an overnight fix.”

Site owners hung on Google’s every word and waited for months and months - while investing in outside support and overhauling their sites - for any semblance of a change that would help them recover.

It’s been almost 2 years, and very few HCU sites have seen any meaningful recovery, despite making dramatic changes and improvements.

This is not the case for websites affected by other algorithm updates outside of the HCU. I’ve personally worked on dozens of recoveries over the years, and there are countless examples of websites who have been able to get back in Google’s good graces after being hit by updates or penalties.

This is not true of the HCU. I’ve been saying since it happened - it feels like HCU sites were “put in jail.” I’ve also done everything in my power to bring visibility to how unfair this is for these site owners, and how these outcomes directly contradicted the guidance Google gave us at the start of the update.

My team and I worked on various HCU-affected sites, with a few great recovery case studies, but many sites were unable to see substantial recoveries. We provided these clients same guidance we provide for all clients impacted by algorithm updates - tried and true methods that have worked for years. These recommendations are always rooted in actually improving websites and making them better for users. I also spent dozens of hours, including weekends, providing free advice to affected site owners simply because I knew how much it was hurting their business, and I wanted to help.

We would have never serviced HCU clients if Google had just been honest from the start - most websites affected by the HCU are going to find it nearly impossible to recover within 2 years, despite making dramatic changes. Again, this is distinct behavior from other Google algorithm updates.

In my opinion, we need to be asking why Google’s communications around this update have changed so much over time, and why these sites seem to have received unfair treatment compared to all other Google updates. I’m a big fan of the work @natejhake has been doing to get to the bottom of these questions. (Although the answers are quite disheartening.)

The entire experience has been devastating for so many site owners, and the inconsistency between what Google said and what Google actually did is something that has forever changed my perspective about the company.

[Matteo Giannone @matteogiannone]

@alessan93669563 This fake will probably steal less money than NKLA.

[Matteo Giannone @matteogiannone]

@InsightGoatLLC @LauraChiocciora @rustybrick Mostly when you decide to add adult content to a national newspaper accessed by millions of teenagers.

[Matteo Giannone @matteogiannone]

@LauraChiocciora @rustybrick I can confirm one of the biggest sports newspaper in Italy had his sportsbetting/gambling hosted section banned from SERPs. They call it partnership, we call it parasite SEO / site reputation abuse. I don't see any action applied to their partner but I expect them to drop.

[Mr Jonathan Jones 🇹🇼🇬🇧 @Jonny_J_]

Google in 1999: No ads, no clutter—just search.

By 2024, Google has evolved into a sprawling empire.

Search is now the distribution engine to AI tools (Gemini, AI Overviews), Cloud services, Gmail, Workspace, Android, YouTube, Maps, Pixel, Nest smart homes, Fitbit, and even quantum computing.

Search Ads dominate, taking over more surfaces, while AI regurgitates content from websites and reshapes how we consume information. Google’s reach extends into how we work, live, shop, and interact, with every query fueling its ad-driven ecosystem.

From simplicity to omnipresence, is this innovation or control? Perhaps both.

[Google Search Central @googlesearchc]

Today we released the November 2024 core update. We'll add it to our ranking release history page in the near future and update when the rollout is complete. For more on core updates:
https://t.co/43pVoYH8k7

[Matteo Giannone @matteogiannone]

@Hobo_Web Geek me !

[Matteo Giannone @matteogiannone]

@Hobo_Web @AnalyticsNinja @KunjalChawhan I'd like to test with my site GSC data (iGaming affiliate site), I see a similar pattern using Semrush data