Remember when everyone thought that Google could live without links?

One day around 2020-2021 I started reading from almost 100% of SEO gurus, that Google was going to fully deprecate the linking part because content and BERT were more than enough to classify the quality of the WWW

Maybe they did not realize that BERT was an AI, consequently an AI will replace another AI.

Now I feel that Google can solely rely on backlinks again and still is holding back because they know that solely relying on of links brings doom.

But content is not any more reliable either for an AI classifier that could be fooled (and is being fooled) with another AI classifier.

Last night I was reading some docs talking about all those n-grams, micro contexts, topical maps, and similar classifying techniques dated around 2022-2023, I could not refrain from grasping a laugh. I've seen a ton of massive sites relying on all this plummet like a rock from a cliff in the last quarter.

This is SEO wheel of life.

Yes, endless cycle.

They admitted it themselves how scary it is.

What required complex infrastructure and millions of $ in the past can now be achieved from a notebook. Not the fastest one either.

tiiberius said:

What required complex infrastructure and millions of $ in the past can now be achieved from a notebook. Not the fastest one either.

Click to expand...

Exactly, I was running an LLaMa-2-7B from my Macbook 2013 last week, and it did work slow but well. Basically, the empire of AI was unlocked.

The problem is that the Google algorithm has now become a clown-fiesta and it is depressing me. I can't seem to pinpoint a single method that actually goes in the right direction. Core updates have switched into a weird rotation system where they null what formerly worked.

Now the trend is to figure out which will be the next map to be playing to adapt the strategy to it.

SirLouen said:

I can't seem to pinpoint a single method that actually goes in the right direction. Core updates have switched into a weird rotation system where they null what formerly worked.

Click to expand...

Topical authority works well and will continue to work because it requires large human input.

AI can attempt to do it but it is limited by what it knows, how it learns and it often makes factual mistakes, misunderstands context or simply just doesn’t honor the prompt.

To build Topical authority, it is not enough just to generate content. You need to create a list of all key phrases, develop a breakdown of these phrases, then a semantic structure.
AI, of course, helps in the whole process. But it is still crucial to match the content to the intent of a given query, it is not enough to generate a wall of text.

As for backlinks - you're right, this is something that makes a difference nowadays, but in addition to this there are behavioral factors.

who's the SEO guru you mention?

Just keep doing it and adapt to changes.
SEO is an ever-changing game and that is what makes it interesting.

Backlinks making the biggest difference nowadays, especially after HCU and spam updates we had last year.

Topical authority is the new buzzword everyone's throwing around
Backlinks will always have it's own relevance

It reminds me of the movie Terminator 2. How the Terminator tries different faces it knows, before finally collapsing into the molten pit. Hoping one of those faces will save him.

Google is doing the same, rotating through different algos hoping to save itself.

SERP quality is deteriorating and Google knows it.

Continue as before and adjust to circumstances.What keeps SEO interesting is that it's a constantly evolving game.

I suspect google employees throwing us off by creating fake articles and fake youtubers about nonsense

And to think that a significant portion of the SEO sphere was downplaying backlinks in favor of content only.

It goes to show that SEO is a never-ending catch-and-mouse game, and we're not the cat.

- You can rank via ctr manipulation, if you competition is low
- You can rank via hundred of different low level backlinks, if your competitors are at the same level
- You can rank via few d0follow dr/da 80-90 if your competition is high. Links from those high auth is like "hey G, I'm Donald Trump(high authority) I vouch for this site or.. I'm Gordon Ramsay, I vouch for this restaurant(high auth + related)".
[N0follow, sponsored tag =/= vouch.]


Regardless content quality etc, you still need vouch/backlinks.

After the vouch > user visit your site by clicking the link(CTR) > G will start analyse each of user behaviour for the exact confirmation.
At first, G will start with lower comp keywords. Before slowly rising to the top keyword, if the behaviour gave positive signal.

This only apply for highly competitive keyword.
For lower comp keyword, the flow will be much simpler, as sending fake ctr clicks or publish high quality content(if your site has lower auth, else instant page 1).

HustleTong said:

Topical authority is the new buzzword everyone's throwing around
Backlinks will always have it's own relevance

Click to expand...

Definitely. I'm not sure, but I think that Koray somehow popularized it?

Nowadays, you read into any online group, and you will see people suggesting that this existed before sliced bread, but seriously, I truly believe that is a buzzword perfectly tailored for gurus, to make those massive topical maps.

I'm researching on this, but the whole ton of massive work that it requires, and the massive decay that most pages that have been profiting from this, and plummeted like a rock from September to December, can be counted in thousands.

The fun part is that I'm not even saying that links are useful at all, because I don't even think so. Today they are useful, tomorrow they are not, and 3 days after they are useful again.

This is the first time in my life I can say that (Google) SEO is dying, in the sense, that the uncertainty of what works at what doesn't is absolute. I don't feel any trust in anyone talking about SEO nowadays.

Yesterday I was completely blown away, with one page ranking top 1, without authority and anything, poor content poor everything and not even black hat in any sense, like pure amount of basic stuff, vs a massive site that just for that single keyword, had like 4 or 5 ultra authority links incoming (say like 80-90 DR links, two like top authority in the niche $3-5K links no joke). If I owned that page, I would jump off a cliff: if that did not work to rank (over a fricking crap page), what would? (and content, technical SEO, and everything was in order also). Even the shit page did not have a glimpse of topical mapping or whatsoever, just a regular ecommerce page without any feats at all.

Given how unstable it's the situation, I feel that the sooner than the later, someone will find a loophole in the most stupid thing ever in the past 20 years (like ultra-blasting full tier 1 or whatever).

Personally, I'm daily tempted to push the red button and throw BH rockets to all my sites, from the lowest of the quality of CTR, mass visits to ultra-blasts, etc., anything that comes to my mind all at once. I don't even think that Mr. G. can issue a good manual action at this point.

I'm so negative because of all of this, that it's already affecting other aspects of my life, I have not barely written in the forum for the past 2 months related to this as well. For me, it feels like the SEO dooms day is approaching.

Everyone is way overthinking it. Google said they are using new AI for the helpful content update. At same time they rolled out Bard, fully demonstrating that their AI is infact retarded. So you have to think "will this content help a retarded AI?" Then write. Smh it's so simple.

Totally scared when heard about the information.

Be calm and go with the flow. Adopt to the changes.

SirLouen said:

One day around 2020-2021 I started reading from almost 100% of SEO gurus, that Google was going to fully deprecate the linking part because content and BERT were more than enough to classify the quality of the WWW

Maybe they did not realize that BERT was an AI, consequently an AI will replace another AI.

Now I feel that Google can solely rely on backlinks again and still is holding back because they know that solely relying on of links brings doom.

But content is not any more reliable either for an AI classifier that could be fooled (and is being fooled) with another AI classifier.

Last night I was reading some docs talking about all those n-grams, micro contexts, topical maps, and similar classifying techniques dated around 2022-2023, I could not refrain from grasping a laugh. I've seen a ton of massive sites relying on all this plummet like a rock from a cliff in the last quarter.

This is SEO wheel of life.

Click to expand...

Top ranks are left for behemoth brands in almost every vertical, and what's left for the smaller fish is to get them links. Google can sell their "content-first" mantra to noobs all they want, fact remains that PageRank is still (comparatively) the most effective scoring system they've got.

Links have been and will continue to be the most important ranking factor.
The problem is that Google devalues a very large percentage of links placed incorrectly in its opinion.
Yes, AI helps it a lot with that. Without AI, evaluating link placement was much more difficult.

What most people don't realize when looking at these leaps in SEO is that the amount of information and domains themselves on the internet has increased many times over, which is why search results have gotten worse. Handling 10 million sites 10 years ago and 12 billion now is a different task.

I always thought that in the current situation, an instant and eternal ban of copies, any copies, would solve the problem for a while. But Google is mumbling, they are afraid to remove copies, citing that there are words in the copy that may carry other useful information.
I, and other developers like me, know this, we know Google's word weights, we know which niches have certain weights and are not to be blocked for copy and we use this to flood the internet with more copy that ranks at the top even though it doesn't carry anything new or useful. And so it will continue to be until Google makes the willful decision to remove any duplicate information permanently.

As Google's bases are filled with garbage, it removes useful old content that it can no longer afford to keep, but keeps this garbage modern.


They talked about thematic authority, no one popularized it, it is one of the sub branches of the BERT algorithm, who is seriously involved in mathematics, he knows that it is thematic authority and the division of phrases into groups of topics determines the links and weight when searching for textual information.
This is the basis, Turbek just voiced it in his works, and the leeches of marketers forced him to start doing what he is doing now. I think he himself is not happy that he got in touch with them, they only know how to make money on the skills of others, while doing nothing themselves (although, the ability to make money on others, apparently also useful and cool skill).

Yes, about thematic authority, it is mandatory, it is determined by which link and what weight will pass to your site. If you ask BARD and ask to give you types of sites from where to put links to your site, he will tell you about thematic authority and show examples of how and where you maximize the effectiveness of placing links to your site. However, if your site is just a mishmash of words and topics, even BARD won't be able to figure out what your site is and give you any good advice.

Also, when you create sections on your sites, topic authority is a must, it is the basis for calculating textual relevance, the more accurate and extensive you take away the topic word map, the higher the weight your section will have on the site as a whole. The more correctly you link it to related sections, the better your entire corpus of topics will be ranked on the site. You will need, for example, 3 good links to get the top 1-3 on a competitive query, and who has a site built worse, will need 50-100 links of the same level.



If you do not use risky technologies, place links only on well-planned and successful sites that are suitable for you to promote, you will not have any jumps, you will always stand evenly in the output, minimal fluctuations will be caused only by various tests of modules in the output, but the core algorithm itself will not affect you in any way.

I have a group of such sites, they are not affected by any updates, they always either grow, or stand in one place having exhausted all queries in the niche. But they can not get a lot of income, because of this policy. Therefore, to make money, you need to use completely different technologies and lose sites and ratings, and again to gain and raise them.