Are your T1/T2 built in SER? I stopped using anything from SER a few month back as the target money pages I used it on either failed to rise or fell in the SERPs.
Pretty sure the time of autogenerated content and mass automation are over. Inboxed you a link to a post I made over on BHW with some screenshots of my analytics over the past few month.
I know how extensively you've tested with SER, and how you have many link sources. Which set-up of yours do you mean? ANY/ALL T1/T2 links from GSA-SER?
"Pretty sure the time of autogenerated content and mass automation are over. "
Why? I'd say it's more like the time of detectable autogenerated content and mass automation is now over.
Throwing buckets of links at a target the way it had been done is definitely not the answer anymore.
Take some of the autogenerated content you think is not detectable and run it through Grammarly just for its spelling and grammar check and see how many errors it pulls up. If they can get that many, I would bet Google can get more.
If you have the accounts required run that same content through the Grammarly and Copyscape duplicate content checker and see if you get any matches. If you do remember that their crawlers are not as good as the one Google use and that Google also have access to all of the pages they have ever crawled to compare the content against. Then there's also rank brain getting smarter by the week.
Regarding the mass automation side of it, my tests this year have suggested that a domain that goes offline a bunch has any potential link juice from it disregarded. I'm guessing this is what the past iterations of Penguin used to do each time Google manually rolled it out but its real-time now. When Penguin 4 went real time it took about a month for my stuff to start dropping in the SERPs and I'm guessing that was the time it took for Google to work out if a link from a domain should pass link juice or not.
You can test your own link uptime by making a random project in SER and letting it build say 10,000 links. Duplicate -> everything on that project 9 times and then reverify one of those projects each day over the span of 10 days. The links successfully re-verified each day will change as links are deleted/domains go offline/domains are rolled back. Sometimes the number will go up when compared to the previous day as a domain has came back online.
I always knew it would come to this point with AI and SE'es.
"...see how many errors it pulls up. If they can get that many, I would bet Google can get more..."
Doubtless, true.
But also, @shaun, please consider this: You are clearly an uber-skilled and articulate writer. You understand language(s).
The average Joe does not write like you. And, there are tons of people writing in a language as a second, or even third, language online, who many not do so quite so skillfully as others.
So, Google has to allow for even excessive errors I would be inclined to consider...
"Google also have access to all of the pages they have ever crawled to compare the content against"
This is the part that's spooky.
I mean, even when I would do SEO by hand, un-spun, after a while, profiles, comments, etc., begin to overlap when promoting with even a larger set of keywords. Could even that now be found to be suspect by Google?!
Where is there room for legitimate promotion? "Getting the brand out" is the oldest adage. Unless I buy ad space, it's not OK?
"The thing with autogenerated content is the more readable you have the settings, the more duplicate content hits you get."
And, the more you keep it unique, and you end up with Alzheimer's word salad, I'm sure that's detected as well on the other end of the spectrum, also.
"Any chance you could paste an autogenerated article into this thread that you think is undetectable?"
Considering finite pool of content...limited number of spinner APIs....considering the reality of your charts...
Probably can't.
But a spinner with custom thesaurus that is far more inclusive of a language's total words and phrases, and better grammar rules, and more concatenating of multiple articles, why not?
VMart
Natural SEO
Comments
Pretty sure the time of autogenerated content and mass automation are over. Inboxed you a link to a post I made over on BHW with some screenshots of my analytics over the past few month.
"I stopped using anything from SER..."
I know how extensively you've tested with SER, and how you have many link sources. Which set-up of yours do you mean? ANY/ALL T1/T2 links from GSA-SER?
"Pretty sure the time of autogenerated content and mass automation are over. "
Why? I'd say it's more like the time of detectable autogenerated content and mass automation is now over.
Throwing buckets of links at a target the way it had been done is definitely not the answer anymore.
I will def check out your analyses on BHW.
Take some of the autogenerated content you think is not detectable and run it through Grammarly just for its spelling and grammar check and see how many errors it pulls up. If they can get that many, I would bet Google can get more.
If you have the accounts required run that same content through the Grammarly and Copyscape duplicate content checker and see if you get any matches. If you do remember that their crawlers are not as good as the one Google use and that Google also have access to all of the pages they have ever crawled to compare the content against. Then there's also rank brain getting smarter by the week.
Regarding the mass automation side of it, my tests this year have suggested that a domain that goes offline a bunch has any potential link juice from it disregarded. I'm guessing this is what the past iterations of Penguin used to do each time Google manually rolled it out but its real-time now. When Penguin 4 went real time it took about a month for my stuff to start dropping in the SERPs and I'm guessing that was the time it took for Google to work out if a link from a domain should pass link juice or not.
You can test your own link uptime by making a random project in SER and letting it build say 10,000 links. Duplicate -> everything on that project 9 times and then reverify one of those projects each day over the span of 10 days. The links successfully re-verified each day will change as links are deleted/domains go offline/domains are rolled back. Sometimes the number will go up when compared to the previous day as a domain has came back online.
I always knew it would come to this point with AI and SE'es.
"...see how many errors it pulls up. If they can get that many, I would bet Google can get more..."
Doubtless, true.
But also, @shaun, please consider this: You are clearly an uber-skilled and articulate writer. You understand language(s).
The average Joe does not write like you. And, there are tons of people writing in a language as a second, or even third, language online, who many not do so quite so skillfully as others.
So, Google has to allow for even excessive errors I would be inclined to consider...
"Google also have access to all of the pages they have ever crawled to compare the content against"
This is the part that's spooky.
I mean, even when I would do SEO by hand, un-spun, after a while, profiles, comments, etc., begin to overlap when promoting with even a larger set of keywords. Could even that now be found to be suspect by Google?!
Where is there room for legitimate promotion? "Getting the brand out" is the oldest adage. Unless I buy ad space, it's not OK?
"Google has to allow for even excessive errors I would be inclined to consider"
But they don't have to rank/index/allow link juice from it.
The thing with autogenerated content is the more readable you have the settings, the more duplicate content hits you get.
Any chance you could paste an autogenerated article into this thread that you think is undetectable?
And, the more you keep it unique, and you end up with Alzheimer's word salad, I'm sure that's detected as well on the other end of the spectrum, also.
"Any chance you could paste an autogenerated article into this thread that you think is undetectable?"
Considering finite pool of content...limited number of spinner APIs....considering the reality of your charts...
Probably can't.
But a spinner with custom thesaurus that is far more inclusive of a language's total words and phrases, and better grammar rules, and more concatenating of multiple articles, why not?