He is one of the guys who make PAA question/answer Spam sites. And here is his Process.

1 article = 10 sites. 5000 articles might be 5-10k different sites. That's why I'm a running a 24 core CPU to make 1 site and it still takes 2 weeks.
What about the keyword research? How do you do that?
 
gpt3 has no factual accuracy.It has good English but my solution does to.

Yes, especially stuff like local SEO. "plumber in [10k US cities".

1 article = 10 sites. 5000 articles might be 5-10k different sites. That's why I'm a running a 24 core CPU to make 1 site and it still takes 2 weeks.
At present nearly all of the AI writers do not have any fact checking mechanism. Only AI models that have human brain capabilities may be able to do that and we are getting there - https://www.tomshardware.com/news/china-builds-brain-scale-ai-model-using-exaflops-supercomputer
 
Ripping PAA directly isn't going to really do it unless the trick is pure volume. I wrote something in a few hours that can rip and turn PAA into tons of pages, but the trick would be either pure volume or PBNs. Plus the DMCA could be an issue.

I do like the rephrasing method so that you're technically unique. Kinda keeps you out of hot water and then you can strive for pure volume. However, I guess part of the trade secret is knowing which data to grab to rewrite to make solid pages. That and how to to paraphrase progmmatically in your own words. But is pure volume say 10k pages enough without push from a PBN?

I do really enjoy coding blackhat stuff so I may toy with some stuff and see what happens.
 
What about the keyword research? How do you do that?
1. Check the US volume
2. Scrape the SERPs
3. Look at the articles and domains
4. Analyze semantic relevancy and authority of the pages
5. Emphasize SERPs with user generated content
6. Sprinkle in a few other factors and calculate my own keyword difficulty score.
 
@Sartre Do you go for the keywords with lesser competition since it's generated?

Thinking about an earlier post you rephrase PAA for a given topic to create that resulting article. If it's not giving away anything too much, do you just use those 4-5 PAA for a given topic or combine more in some way so that the resulting article is more than say 300 words?
 
@Sartre Do you go for the keywords with lesser competition since it's generated?

Thinking about an earlier post you rephrase PAA for a given topic to create that resulting article. If it's not giving away anything too much, do you just use those 4-5 PAA for a given topic or combine more in some way so that the resulting article is more than say 300 words?
Yeah I definitely go for lesser competition. <30 according to my algo and I prefer keywords with UGC on page 1.

I use 3-15 PAAs per article. Sometimes it's hard to find a lot of relevant stuff.
 
@Sartre - one word comes to mind about that AMA thread and process: epic.
btw, about that VPS optimization article/guide - is it ready? Can we get a heads up here when you'll post it? Thank you!
 
@Sartre - one word comes to mind about that AMA thread and process: epic.
btw, about that VPS optimization article/guide - is it ready? Can we get a heads up here when you'll post it? Thank you!
I wish it was ready, I had a CRAZY amount of work and orders recently. I will give a heads up.

Wait is it even allowed to post a link to an article I write on BHW? I guess I will ask a mod.
 
what tools does he use to create articles ???
 
1. Check the US volume
2. Scrape the SERPs
3. Look at the articles and domains
4. Analyze semantic relevancy and authority of the pages
5. Emphasize SERPs with user generated content
6. Sprinkle in a few other factors and calculate my own keyword difficulty score.
Can you turn this into a tool? Maybe a SaaS? Because it will be insanely valuable and helpful if you want you can charge per keyword rather than a monthly price.
 
Can you turn this into a tool? Maybe a SaaS? Because it will be insanely valuable and helpful if you want you can charge per keyword rather than a monthly price.
Why would he want to make competition for himself
 
Can you turn this into a tool? Maybe a SaaS? Because it will be insanely valuable and helpful if you want you can charge per keyword rather than a monthly price.
not worth it:

1. small money charging for a SaaS
2. have to do customer support
3. make a GUI (huge PITA)

I'd rather run my own sites + make sites for investors.
 
The GUI is time-consuming without a doubt, and it's going to be custom since it's how the client is interfacing. Host-wise Amazon cloud talking to Stripe is a common way that works well and is dependable and scales easily.

However, the GUI is fairly one-time in that once you set it up, you may be tweaking algorithms, but it'll be mostly there without much more to do. The support is where it keeps pulling you back to keep an eye on it. I think freshdesk had some low-cost options for a ticketing system and you might be able to throw them a few bucks to answer the easy ones for you.

So it may be more effort upfront but it might not be that bad once it gets going.
 
Back
Top