Below are the most recent 30 comments. I try to keep it clean of comment spam, but some times things
get through and it takes me several hours to get to it. So please excuse any of that comment spam.
The second test they literally tried a play on the "Invisible text in the footer" trick and are surprised that Google didn't count the invisible text in their index? The second page was a blank page where the schema was not relevant to the content on the page, so it would be ignored.
At least the first test is interesting, but Tokenization isn't the whole story, otherwise LLMs couldn't understand code. If you take code and paste it into the Tokenizer example on ChatGPT, the LLM also "breaks" the code. But we know that models can understand code and respond to it. If you ask questions of an LLM using structured data like formatting, the model interprets it appropriately. https://uploads.disquscdn.com/images/344877f9f249f1595b2c4147b1faff56772863196ea24145134f1200ed16922d.png
As others have mentioned below: We know that Google uses Schema to inform their index. It is not a ranking factor (you won't improve ranking from spot 3 to spot 2 implementing schema) we DO know that schema helps Google better understand your page so you can appear for a broader range of terms and my company has case studies showing that schema can improve the quality score of a page, making it more efficient for ads.
In my own testing, We also know that these models CAN parse schema when fed a page, so they understand the context. It's not a silver bullet and should never be treated as such, but I think that's a far cry from the "Schema is useless" talk, especially when so many people are still doing genuinely useless stuff like keyword rank tracking in AI models (Or worse, using AI for Keyword research) and paying hundreds of dollars a month for it.
Again, Schema isn't a silver bullet. Anyone trying to sell it to you as the "way" to "win AI" is trying to just take your money.
My theory?
Google is doing this, so we can’t get reliable and accurate third-party data. This way, we can only rely on the data Google gives us (i.e, Google Analytics), which they can manipulate and lie to us. Don’t forget, this isn’t the first time Google has attempted this. Earlier this year, Google did ‘something’ to the search results, I believe they added some hidden elements or something, and it made many scraping software and keyword tracking not work as normal.
It is also to deter OpenAI and all the other AI startups from scraping content online conveniently. This way, AI startups will have a tough time scraping for data within their budget. They will have to buy proxies and whatnot, which will skyrocket their expenses. It will benefit Google in the long run since people will just rely on Google for the latest, correct answers. For other words, they're setting barriers to data harvesting.
Sometimes I feel that they didn't just intentionally deteriorate the search results' relevance to push us to Adwords; it also hurts AI companies who rely on Google to scrape for data, since it will be irrelevant and doesn’t answer the question. Reddit always rank first and, coincidentally, Reddit has security measures to block scrapers. Then they put up Youtube, Pinterest, and Facebook which won't be much value for data harvesting.
Anyway, when content creators like us can’t get accurate data to let us know where we’re ranking; Google can keep shuffling the search results crazily behind our back. Then they can manipulate or even delay Analytics data. We will think all is ‘normal’ and we will continue to write content like mules oblivious to Google’s ploy. Then Googlers can still tell us, “SEO is not dead.”, or “Keep writing helpful content.”
<blockquote>google is fucking greedy beyond belief.</blockquote>Greedy they are, but it's not stopping them from pushing their gaslighting PR campaign. Just saw the story below on CNBC with Youtube bragging how much they are paying out. All of the comments I can recall here, from those producing Youtube videos, is them being paid peanuts. But being paid peanuts is still better than what Google search pays creators.
<b>YouTube says it has paid creators more than $100 billion over last 4 years</b>
<blockquote>YouTube said on Tuesday it has paid out over $100 billion to creators, artists and media companies since 2021.</blockquote><blockquote>The milestone comes as the Google-owned platform marks its 20th year and pushes to cement itself as one of the world’s most lucrative media businesses.</blockquote>Full story: <a href="https://www.cnbc.com/2025/09/16/youtube-creators-pay.html">https://www.cnbc.com/2025/09/16/youtube-creators-pay.html</a>
Can someone explain to me why Google would make this change? Honestly, what's the point of removing the top 100 results? And yes, we do fking care about 30+ positions.
maybe.... just maybe.... it will also click on its OWN ads to purchase from advertiser's websites... then they charge advertisers for it, while also taking a cut from the customer's purchase for 'processing fee'.
google is fucking greedy beyond belief.
I think it's only a matter of time before Google uses their unchecked monopoly to scrape our product content, present it to users with their own inflated price so they can forcibly take a slice of the transaction by making consumers pay a higher price. The foundation is being built...
<b>Google launches new protocol for agent-driven purchases</b>
<blockquote>On Tuesday, Google announced a new open protocol for purchases initiated by AI agents — automated software programs that can shop and make decisions on behalf of users — with backing from more than 60 merchants and financial institutions. Called the Agent Payments Protocol (AP2), the system is meant to be interoperable between AI platforms, payment systems and vendors, providing a traceable paper trail for each transaction.</blockquote>Full story: <a href="https://techcrunch.com/2025/09/16/google-launches-new-protocol-for-agent-driven-purchases/">https://techcrunch.com/2025/09/16/google-launches-new-protocol-for-agent-driven-purchases/</a>
Great point, Kevin — I completely agree that pairing GSC data with log file analysis is key to separating genuine user demand from bot noise. I find this perspective very interesting. Have you come across any specific frameworks, case studies, or resources that explain how to effectively combine GSC and log files? Would love to explore this further with some supporting material.
Judge Meathead could have dealt with this in his ruling, but he succumbed to being corrupted and should be removed from the bench. But this speaks to the larger problem of allowing foreign nationals to become judges and lead companies (Microsoft, Alphabet) which are hostile to American freedoms and liberty.
I wonder if anyone has filed a DMCA takedown request for stolen content appearing in AI Overlords? Would be nice if Google was flooded with millions of these DMCA requests and also millions of lawsuits from small publishers who have been severely harmed by the Google Crime Syndicate.
Trump isn't going to help anyone other than big techies. In fact the one who nicknamed Sleepy Joe is now being called Nappy Trump and said to be showing signs of early dementia. Too many Geriatrics in DC which is why age limits are needed.
https://uploads.disquscdn.com/images/71a372baa90d44bb4b788422a63cbe1527463e4bf3b2636889ba60b2b66c59e3.png
<b>‘This Isn’t Normal’: Psychologists Sound Alarm on Trump Dozing Off</b> - <a href="https://www.yahoo.com/news/articles/isn-t-normal-psychologists-sound-170735970.html">https://www.yahoo.com/news/articles/isn-t-normal-psychologists-sound-170735970.html</a>
<blockquote>Two psychologists are warning that President Donald Trump’s frequent confusion and mounting physical issues may be pointing to something more troubling than just aging.
On their podcast Shrinking Trump, psychologists John Gartner and Harry Segal argued that the 79-year-old president is showing both physical and linguistic symptoms of what they call “early dementia.”</blockquote>
Please do. But if you delete it quickly I will probably miss it since I'm not spending as much time online anymore except for work.
I have a feeling we're getting closer to AI Scrode Mode being the default. Ranking #1 now brings no traffic anyway, so going full AI Scrode Mode is the next logical step for the Google Crime Syndicate.
https://uploads.disquscdn.com/images/7dc9d793c86f6e3f128da6fd9105837959ef59445c40124c7b8e269c4f04a32e.png
My pages are not getting indexed. I’ve fixed all technical errors and duplicate content, but Google is neither crawling the site nor indexing it.
David, agreed — schema isn’t a ranking factor. That part’s clear.
But saying it does “nothing” isn’t accurate either. Google explicitly documents its use of structured data to power Merchant Center feeds, Knowledge Graph entities, FAQ/How-to rich results, events, and product attributes. Those aren’t “magic,” they’re retrieval inputs.
And retrieval matters: MAGI-T, AI Overviews, and RAG pipelines can only cite what makes it into that shortlist. Schema doesn’t get you into the model, but it does change the pool of candidates the model can pull from.
There are things I have written from my head (only me) that are in AI overviews from my website. It does not quote me but it is so obscure that it could only come from my brain. And now, they are using my own content to rob me, that is not fair use.
https://uploads.disquscdn.com/images/8d5b66d1f24798bf9cb60e8b7dbb2dd2a1f9ffaa331aca16b4f8a4f7a53a984a.png I am in UK - so maybe I have it before in US, or I am just part of a test pool before it launches?
they basically have covert contract with us, they expect us to continue writing, while not telling us that it's terrible business for us :DD that's why they really feel no need to comment, because we keep writing while they know it's worst idea, but it's ideal for them
but where do you see that button "AI mode"? I don't see it anywhere
https://uploads.disquscdn.com/images/322b2934791955bc41048683457b34f567bde57956b3a57b332f84a320aef0a7.png