Is A.I. Content Causing SEO Indexation Problems?

[CCarter]

psyops is a very real thing in SEO

There is a massive psyops to have website owners publish content written by humans?

Everything discussed is based off of failed experiments of people using A.I. content - basically data. Even the examples give above you can clearly see algorithm hits and death spirals, again more data.

You haven't provided any data to suggest otherwise, just "vague" people "winning" with A.I. content - no data.

Show us data stating we are wrong and we'll gladly make the corrections to the argument. Otherwise it's just speculation that ends up wasting a ton of times for newbies that believe the hype.

 

[glamdring]

There is a massive psyops to have website owners publish content written by humans?

Yes, there is a psyops to say AI generated content does not work / is not effective.

I can't upload an image from Ahrefs, but you could check out askinglot.com

Peak traffic was around 5m. This was already slapped down so there's no issue to use it as an example. There are many other, more refined examples still hiding in plain sight today.

Once people have their setup they can automate everything down to a tee. Pump out however many posts at whatever velocity they like.

 

[Ryuzaki]

I can't upload an image from Ahrefs, but you could check out askinglot.com

I was curious so I snapped a screenshot and uploaded to Imgur:

​

I think we need to define what "works" means. It clearly works in the sense that Google will index it and rank it. It clearly doesn't work because Google will essentially deindex or derank it to zero, too, as soon as they catch on, whether manually or undoubtedly (in the future) algorithmically if it isn't already that way.

But yeah, if you can just flick a switch and throw up dozens more sites that last a year and make $5k, $10k, $20k a month each... well... Anything any algorithm tries to do, by nature, has an exploit. They all have a weakness, even if it's dozens of sub-algorithms working together. Neo is reborn every time Zion is destroyed.

 

[CCarter]

This was already slapped

So you don't have an example.

I don't get it, do you think we are spammers trying to be on hamster wheels every year rebuilding after each crash?

What we concentrate on here is for long-term impacts which result in exits/acquisitions. Think about it, why would a spammer waste our time asking questions on this forum about the slap he got? He wouldn't cause he knows why.

The people asking questions here are in it for the long-haul and dedicate hours and days per week to create their business. This isn't black hat world. The word "Builder" is in the name.

What you showed me is not a business, it's just spam that will be nuked eventually. Like a hustler going from one hustle to another, not a business.

 

[UFO]

They also said that the true mission wasn't to make money, although they can certainly scale the number of sites and niches they're building these sites in. The true mission is to build these on low powered domains to find out which keywords are weak enough to tackle on real sites. It's a research game for this one guy anyways.

This is one of the more intriguing and underlooked side stories of this whole debate, in my opinion.

I recently ran across this Reddit post. This dude that automates PAA scraper sites did an AMA and his whole angle is to fish for zero search volume or low search volume KWs that hit page 1, then have writers create unique content for his white-hat sites. He then deletes the spam articles from the PAA sites so as not to compete with himself. He does not monetize the PAA sites at all, they're purely used for KW research and to make sure his content investment gets the most bang for the buck.

He claims a 99% hit rate for page 1 using this method when creating new content for the white-hat sites based on ranking page 1 with the spam sites. He also states that it took him and a full-time developer a year to automate it and that ZSV KWs bring "pretty ok traffic."

To me, that sounds like a lot of effort for "pretty ok traffic" but clearly others are doing it as well so makes me think that there's some gold in there and that they're on to something. Another part of me thinks that the cost of a full-time dev for a year plus this dude's time cost coding alongside the dev could've been reinvested in content to grow a white-hat site and that the whole automation thing is just adding extra steps.

So the question is, is it worth it to have a 99% hit rate on easy, low-volume KWs? If your white-hat site is authoritative enough, you should be able to rank for those with a high hit rate, and higher traffic terms anyway. Why not just focus all effort on the white-hat site and grow that instead of fucking around trying to suck up crumbs? GSC would be feeding you more ideas as you go too based on the current content and relevance instead of having a separate spam site doing that.

I guess once you code the thing, it's pretty much done and can be reused at scale. As someone that cannot code, having something built like this seems like a huge mountain to climb. I'm sure some of the coders here can tell me if this is just plain stupid or if it has some merit. Interesting angle though, nonetheless.

 

[glamdring]

You said AI content doesn't work well and I showed you it does - the result is above (Askinglot screenshot). Now you are deviating from the initial messaging.

How far you want to go with AI content is up to you - that is the point. Some, like Askinglot will spam. However, there are many other sites going far beyond where you probably wouldn't even know you are reading AI content.

There are people mass building AI sites so they can take the very real keywords with volume, and then move them into their money sites. The possibilities are endless but you closed the book on something when it's not black and white.
----------

AI is actually a prime candidate for the Avalanche method discussed in here.

I don't get it, do you think we are spammers trying to be on hamster wheels every year rebuilding after each crash?

Spinning round on a hamster wheel? That is already each and every one of us. Google releases a Core Update and even if you did everything by the book you can still get hurt. Then you adapt and rebuild if you need to.

 

[freshpeppermint]

How old is the site(s) having the issue?

Started it in December.
The thing is, I believe after blog commenting, quora, and other things I built those do follow DR 30-50+ links on websites, like where you sign up on this website and get links built this website I believe. As well as some garbage put domain and auto-generate backlinks like this https://www.backlinkr.net/. I did this because I would expect google to not look into this or slow a website down because otherwise people can just do this for competitor's websites and get them deranked.

Turns out that the article that I instant indexed didn't seem to rank. So what I am going to do now is maybe try to write a brand new fresh article rather than rewriting and updating the AI-written one.

I didn't do any of these things except AI and instant indexing on my other site and the strange thing is that MAJORITY of the articles like 99% are AI on this site yet it gets hit then recovers.

 

[CCarter]

That is already each and every one of us.

If you are selling e-commerce products through your brand you can't just keep creating new websites cause your blackat SEO technique stops working.

I believe you are thinking of SEO spamming versus owning a business. I'm pretty clear what I am referring to here.

Your model is not a business, it's just spam.

 

[DanielS]

Started it in December.
The thing is, I believe after blog commenting, quora, and other things I built those do follow DR 30-50+ links on websites, like where you sign up on this website and get links built this website I believe. As well as some garbage put domain and auto-generate backlinks like this https://www.backlinkr.net/. I did this because I would expect google to not look into this or slow a website down because otherwise people can just do this for competitor's websites and get them deranked.

Turns out that the article that I instant indexed didn't seem to rank. So what I am going to do now is maybe try to write a brand new fresh article rather than rewriting and updating the AI-written one.

I didn't do any of these things except AI and instant indexing on my other site and the strange thing is that MAJORITY of the articles like 99% are AI on this site yet it gets hit then recovers.

So...

1. Your website is 6 months old (sandbox)
2. Your content is AI garbage (just admit it- it is)
3. Your website has no authoritative backlinks (autogenerated spam links don't count)

Take a look at those 3 things. You know you've been screwing yourself- you've been here long enough to know this. And now, Google is showing you this. Posts are getting deindexed (as they should) and instant indexing isn't even working (stop doing this).

I don't even know what I'd do at this point. Is this also the website that you purchased the DA/DR increasing Fiverr gig for? If it is, yikes.

I tried AI content with Jarvis (Jasper) and I was incredibly unimpressed with the output. So, I can't even imagine what the content on your site is looking like.

 

[Ryuzaki]

3 reasons all you AI content SEO's are doomed from the start if you can't scale to the moon right now:

1. Google is getting smarter

Google is getting better at detecting spammy content and penalizing websites that try to game their algorithms. This means that if you want to rank well in Google, you'll need to create quality content that's relevant to your audience.

2. Content is king

Content is still king. There's no denying that. And while some businesses are using AI to generate content, they aren't doing it nearly enough. They should be generating content automatically and making sure that it's optimized for SEO.

3. The future of SEO is machine learning

Machine learning is going to change everything about SEO. It's already changing how we write content. We're seeing AI take over many different tasks that used to require human labor. So what does this mean for SEO? Well, I think we're going to see SEOs being replaced by machines.

--------

The top line is a prompt I fed the free Frase.io demo. The rest is what it spit out. They were on point at first and then immediately derailed by point #2. "AI sucks" quickly became "you should use way more AI content", then by #3 it wasn't even on topic any more.

I know you can keep it "on the rails" by feeding it an outline of several headers or prompts. But... eh.

I'm not on the up-and-up about where to test all this out, but here's one I found from the MIT-IBM Watson AI lab and Harvard NLP group:

If the text is in green, it makes sense and is a word that would be expected to appear in the flow of text based on the context of what is to the left. If it's yellow it's in the top 100 words that might appear based on the context. Red means top 1000. Purple is "nah bro".

So, having MORE green and no yellow, red, or purple is bad. It indicates that a really crappy thesaurus or whatever is being used and the complexity of human speech and writing isn't present.

Since there's some jargon in here, I'd say this passed THIS test with flying colors. But any human will read this and be like "wtf" nearly immediately.

--------

Do I have some kind of point? No. I just wanted to play with it and figured I'd share the results.

For the type of content I publish, by the time I generated an outline for this thing to flesh out, and then read it to make sure it's not completely bonkers, I could have written it and known it was fine. Or paid another human to do it.

At the end of the day, it's apples to oranges. No serious website is publishing this crap unless they're doing financial reports or whatever with ad-libbed data. If you're a spammer, have fun fighting the war many of us waged over a decade ago before seeing that the grass truly is greener on the other side.

My recommendation is that if you're new and want to be a spammer, go for it, BUT go ahead and start a real project too. Do it now, and add content to it every month and just make it a habit. 5 years from now when you're sick of spamming and starting over and spinning your wheels, you'll have a nice, legit, white hat, human written website that's already 5 years old, getting traffic, and making money to start off with. Rather than starting at ground zero after being in the game for 5 years.

I said this same thing to every moron on Wickedfire and Penguin came and wiped them out and they all quit and become cryptotards instead and are wiped out AGAIN. Just have one single legitimate project growing in the background. You'll thank yourself.

 

[CCarter]

At the end of the day I posted this on "BuilderSociety" and not BlackHatWorld. In the OP it states:

"Assets which they are staking their future on."

"Assets which they cherish and are suppose to give them financial freedom."

All the AMAs on this forum have been about successful exits, 6-7 figure exits. The Digital Crash Course is literally revolving around creating online assets.

A website that gets destroyed in 3 months or 18 months isn't a business. It spam, fine, but that's not what we are HERE for.

It would be like me walking into an igloo in Alaska and asking why everyone is wearing coats when it's hot as hell in Florida. "Cause that's not where we are at right now."

I wouldn't be giving advice to "explore different marketing channels" and "increase social presences" to assets that aren't going to last more than 3-18 months. That would be a complete waste of your time.

You can't do a 6-7 figure exit with this A.I. crap. And if by some miracle you pull it off and then the new owner gets Google slapped, expect a civil lawsuit coming your way for fraud from the new owner. You better have your attorneys' retainer account filled up.

We're here for longevity. That's implied in name of the forum and all the advice given here.

 

[Boy]

Something that nobody is bringing up either is monetization.

Of the AI sites I track, none, and I mean not a single fucking one is monetized with a decent AD provider - No Google Adsense, no Ezoic, no Mediavine, no Adthrive; This tells me these folks are also wise to what AI folks are doing.

If you can't even monetize properly with a decent display ads provider, what the fuck is the point? I guess they're happy with the low tier crappy ads that fuck with UX and make pennies?

Even though they have tons of estimated traffic on tools like SEMRush or Ahrefs, they aren't monetizing in a way that even makes it worthwhile. That or if they reach a certain level of traffic, they get manually reviewed and tank - or this new algo update has worked and these folks have tanked because of that.

As the saying goes "You can only polish a turd so much - that shit is still a turd."

Even a $2 RPM on a site getting 2 million visitors a month where all you had to do was turn a script on isn't terrible, especially when you can do that 10, 20, or 100 times a day just buying domain names.

Yeah, it's a crash-and-burn operation, but let's not pretend the money isn't bad compared to the effort going in. All the big examples around various communities are these big boys that are doing 2, 3, 4, 5 million estimated traffic, but we never hear about all the medium sites doing 2, 3, 4, 5 hundred thousand a month in traffic. And again, highly unlikely it's an operation's only site.

Yes it's spam, yes it's stupid, but shit, it makes sense that it's the exploit of the month.

 

[CCarter]

Guys, the reason I tell you it's a waste of time it because most of you cannot pull it off.

I know I can pull it off. I believe Ryuzaki, grind, and eliquid could pull it off, but that's not to brag, it's cause we've been doing this for a LONG time.

If you Google "site:wickedfire.com" and my username and "markov" you'll see a post where I give out a free script to automate article posting and using some "voodoo" on how to make each article unique. Thats a post from 2008.

We're not new to the darker arts of SEO. If you were to go through my posts on WickedFire, or if you can find old posts from BlackHatunderground, there is literally a ton of blackhat techniques talked about and explained with examples. Shit we got Elon Musk crying about twitter bots now, ya'll don't remember the traffic leaks "twitter bot army" post from nearly a decade ago? (there is a copy of it in the crash course). My bad little brother Elon.

Here is the timeline on for @freshpeppermint:

In the beginning of December I told him it was a complete waste of time here: Are AI content writing software/services any good

He rebutted, and I quote:

I am gonna prove Jarvis works. The articles are honestly good just requires a bit more effort.

Fast forward nearly 7 months later he stated he was un-successful.

Could he have pulled it off - in theory, yes. But he didn't possess the skills to or really the experience.

And the people that I would put money on COULD pull it off are telling you it's a waste of time. Don't you guys think that if this was not a waste of time one of the ancients would be doing this non-stop?

At the end of the day, Mr Peppermint wasted 7 months of his life going down some zip-zagging route cause "someone somewhere on the internet" sold him Jarvis and this A.I. route was the way to go.

Any that's what you are doing now @Boy. Giving false hope that there is some "secret formula/combination" that's going to get these noobs to the promised land where you rank with A.I. content and generate $100K+ monthly revenue for short periods of time.

None of us gain anything by telling you guys it's a waste of time, except hopefully some of you guys listen and don't sink months or years of your life doing some nonsense. If Mr Peppermint listened, what would I gain? Nothing.

But what about Mr Peppermint?

What could have Mr Peppermint done in 7 months? If he was posting an article every day or every other day, he would have 105 to 210 good solid articles on his site that have aged and are gaining traction.

Instead he's fucking getting de-indexed and fucking up his site's stats in the eyes of Google. Sounds like a complete waste of time.

 

[voLdie]

IDK why but you saying "mr. peppermint" made me lol.

 

[Zyzz]

Something that nobody is bringing up either is monetization.

Of the AI sites I track, none, and I mean not a single fucking one is monetized with a decent AD provider - No Google Adsense, no Ezoic, no Mediavine, no Adthrive; This tells me these folks are also wise to what AI folks are doing.

If you can't even monetize properly with a decent display ads provider, what the fuck is the point? I guess they're happy with the low tier crappy ads that fuck with UX and make pennies?

Even though they have tons of estimated traffic on tools like SEMRush or Ahrefs, they aren't monetizing in a way that even makes it worthwhile. That or if they reach a certain level of traffic, they get manually reviewed and tank - or this new algo update has worked and these folks have tanked because of that.

As the saying goes "You can only polish a turd so much - that shit is still a turd."

I have seen a good chunk of them monetized with GCPPs like adpushup, monetizemore, and the likes...

 

[Trump]

Guys, the reason I tell you it's a waste of time it because most of you cannot pull it off.

I know I can pull it off.

Lmao

 

[Boy]

Any that's what you are doing now @Boy. Giving false hope that there is some "secret formula/combination" that's going to get these noobs to the promised land where you rank with A.I. content and generate $100K+ monthly revenue for short periods of time.

When?

 

[ItsYoBuddy]

I'm not even close to the level most people here are at so maybe my opinion doesn't matter, but I was always assuming that most people here don't use AI content. I still work a regular job but I still manage to spend money on writers from upwork every week. For me, I'm just trying to build for the long term and adjust as much as I can when updates happen. I lost a bunch of traffic to my one site, but I was being an idiot. I was jumping from topic to topic and not focusing and dominating on one topic. Lesson learned and making the adjustments.

 

[bernard]

I've used AI content for low effort linkbuiling with PBNs. They seem to be 50-50 getting indexed or not. Mostly it's a question of the authority of the site they're published on.

One thing I remember to do is make it unique enough, by being sure to write good titles and have some unique subheadings when possible.

As in, decide the content in advance, don't let the AI do that. Just let the AI write what you specify in the subheadings.

Overall, I am moving away from AI writing for even this.

It's sort of useful in niche situations such as writing a blog post for PBN or site in a different language. It's pretty as a translator too.

 

[Ababilstar]

Someone has a real good troll account they launched onto BuSo, that's all I gotta say.

Good luck.

Each time I see @freshpeppermint talking, I remember this. And I believe this strongly.

 

[Samwise89]

It’s interesting to see some of the justifications going on. I imagine it was the same for link spam. And when Google wipes it out they’ll be called “evil”.

It’s pretty simple from where I’m sitting. Either they stop it and essentially save the internet, or they don’t stop it and ruin their own business. Pretty obvious what’s happening and going to happen.

 

[wikibum]

I was thinking of testing this AI stuff on a separate blog but no matter what angle I look at it from, it doesn't seem like a good investment AT ALL.

The most obvious reason is that it's super easy for anyone with some coding knowledge to reverse engineer their crap. It would be so easy to spot the writing even if someone edits it after its written.

Mainly because people who use it will only edit 10% of the entire generated text. If an AI user ends up editing 50%+ of the generated text, then what is the point of using the AI. Might as well write up a new article at that point. Plus, editing an article takes much longer than creating a new one (at least for me)

 

[Ryuzaki]

My LoF partner predicted this was coming a couple weeks ago and here we are:

Source for Image above. RankRanger, Semrush, and seoClarity are all showing a massive drop (as high as 50%) in the number of SERPs that are showing PAA.

PAA is People Also Ask, which are the questions that Google lists in the SERPs that the AI content people are scraping and using to generate articles full of questions and short answers, specifically to appear in the PAA section.

Soooo, it definitely appears that Google is taking action against AI content, from not indexing it to removing the source they're using to generate and game the search engine. The attack and death of AI content isn't coming, it's already here and being actively pursued.

You know how they stopped punishing and penalizing for bad links and just started ignoring them and letting you waste your time? Same concept here. Rather than let you reap the benefits for a while and then be deindexed or deranked, they're now just not indexing you. Pretty smart approach, all the way back to basics. No indexy, no ranky, no money.

 

[CCarter]

People Also Ask

I confirmed that is true. Ironically I opened the backend up this morning of SERPWoo and saw this "725" code dropped dramatically starting several days ago, I was going to investigate later on today. However after seeing Ryu's post I went int to confirm - and yes indeed "725" corresponds with our PAA code's execution:

They're up to something! That is about a 50% drop.

 

[Grind]

There is only 1 person that can help you now.

Ranking 100% AI content page 1, don't ask me anything.

They're up to something! That is about a 50% drop.

The new leverage points this opens up are fantastic.