[Journey] 1 million UVs/month in 12 months using AI generated content. Let's do it!

[xczak]

The site is completely down. Did you notice? What could be the issue?

It's not down, just your region is blocked by Cloudflare settings, you can see it through VPN.

 

[prvncqr]

It's not down, just your region is blocked by Cloudflare settings, you can see it through VPN.

No, I mean the site ranking has dropped significantly.

 

[Tlermear]

Looks interesting, following and good luck!

 

[GVald]

Looks like it's going well, I'm also like 90% done with my bot, it's taking between 400 and 700 seconds to produce a full article with relevant subheadings etc.. I plan to run the whole thing on Colab Pro+ see how it does because my computer is nowhere near strong to those machines.

 

[moneyflower]

Looks like it's going well, I'm also like 90% done with my bot, it's taking between 400 and 700 seconds to produce a full article with relevant subheadings etc.. I plan to run the whole thing on Colab Pro+ see how it does because my computer is nowhere near strong to those machines.

Please keep us updated on wether or not Colab Pro+ works well for it.

 

[prvncqr]

Can you please give me a reference to installing the correct version of pytorch for using pegasus with cuda?

 

[Sartre]

Hi bro, also curious if u do a competition check for each keyword you load in?
for the website u using the bot, are all articles created in the same niche?

thanks, bro

yes to both

am also very confused of this part, bro,
how can you add "semantically relevant keywords"? where do these relevant keywords come from?
you add them in the kws scraping list? or scraped from the result page?

Thanks so much if u can give a bit of explanation.

they come from my database, I scrape the serps for each keyword and top 10 articles, find the most popular n-grams in each article to extract the most relevant phrases for each keyword. My app is over 10k lines of code at this point, after refactoring and optimizing it twice.

i checked serp api sites but they are more expensive than you said. and limitation is on the limit. 5.000 search 50 usd how did you do it with much cheaper ?
1 search means 1 keyword right ? then just 5.000 keyword you need to pay 50 usd. i saw another thread guy had 400-500k long tail keyword thats end up more than 5000 - 6000 usd

you have to dig deeper. there are cheaper options, and also I only check keywords with volume.

No, I mean the site ranking has dropped significantly.

they got hit by manual action, traffic almost down to 0.

My sites are still doing fine, they are not as spammy. I'm working now to make sure that they will pass manual reviews.

Unfortunately I was sick(almost ended up in the hospital) for almost 7 days, but yeah, things are still going very fine:

 

[Shaybib]

extract the most relevant phrases for each keyword.

How do you check relevance?

 

[Daemon]

Check these:

mvorganizing.org
dengenchronicles.com
sidmartinbio.org
easierwithpractice.com
janetpanic.com
sluiceartfair.com
idswater.com
colors-newyork.com
holidaymountainmusic.com
restaurantnorman.com
greedhead.net
rehabilitationrobotics.net
everythingwhat.com
askinglot.com

people are banking HARD on AI-generated sites... I hope I'm not too late to the game before Google penalizes this.

The first site got to 5mil visits/month in 6 months

I think I've ended up at a lot of these types of sites recently via Google, I would never have thought they were auto-generated.

Most of these seem to still be doing well according to SEMRush.

A number of these sites seem vaguely similar in terms of layout and structure. Are they using some shared premium autogenerated content CMS or software? Or do each of these sites have their own custom software?

 

[Sartre]

How do you check relevance?

extract n-grams from top 10 articles in the SERPs for each keyword

 

[haykuroo]

Firstly, thanks for such an awesome read. You’ve inspired me to take action and build my own tools. I have a question for you about keyword research. If we’re aiming for tens of thousands of keywords, how do we properly research those? Usually we might look at

• Backlink/Referring Domain count
• Allintitle results
• Relevance
• Content length and quality
• Types of websites ranking
• SERP History

But how to assess all these metrics for such a high volume of keywords? Perhaps we just look at search volume using Moz? Not expecting you to divulge too much info on your strategy, just some pointers on doing research for at this scale? Thanks.

 

[Shaybib]

extract n-grams from top 10 articles in the SERPs for each keyword

You extracet the most important phrases from the articles, but how do you check that the phrases are semantically related to the keyword?

 

[Lioncern]

I will like to follow journey because I'm doing a similar project. Content quality is important.

 

[Sartre]

You extracet the most important phrases from the articles, but how do you check that the phrases are semantically related to the keyword?

data science. I'm sorry but it's too long to explain statistics here. You need to interpolate the data between keywords.

 

[Sartre]

Firstly, thanks for such an awesome read. You’ve inspired me to take action and build my own tools. I have a question for you about keyword research. If we’re aiming for tens of thousands of keywords, how do we properly research those? Usually we might look at

• Backlink/Referring Domain count
• Allintitle results
• Relevance
• Content length and quality
• Types of websites ranking
• SERP History

But how to assess all these metrics for such a high volume of keywords? Perhaps we just look at search volume using Moz? Not expecting you to divulge too much info on your strategy, just some pointers on doing research for at this scale? Thanks.

google SERP apis

 

[haykuroo]

google SERP apis

Sorry, to be clear, I’m asking about selecting keywords, not getting search engine results.

 

[Sartre]

You extracet the most important phrases from the articles, but how do you check that the phrases are semantically related to the keyword?

let me show you an example. 2 keywords that contain the word "rust". one is about the programming language, the other is about a nike sneaker

Sorry, to be clear, I’m asking about selecting keywords, not getting search engine results.

I'm scraping PAAs and Autocomplete for the most part. If that's what you're asking.

 

[ProfitNow]

That is a big target for UV in a month, i will keep following this journey thread.

 

[seowarrior008]

they come from my database, I scrape the serps for each keyword and top 10 articles, find the most popular n-grams in each article to extract the most relevant phrases for each keyword. My app is over 10k lines of code at this point, after refactoring and optimizing it twice.

Thanks for the reply, bro.
If you do a new article in this way, how can you set up the h2 headings in new article? Just simply use those related keywords? or scrape from the original article?

If you scrape the content for each kw, including main kw and related kws, how does ur bot know what kws to be used for a new article, then move on to the next article creation? since you use multi kws scraping for one new article creation, don't you?

thanks a lot if u can answer these.

 

[GVald]

Looks like it's going well, hope all is good concerning your health @Sartre. I also finished writing my script, it basically does everything (I only have to set up the site and do kw research), but I'm running it from Cloud Computing, since I'm using GPT, not exactly the same method as you, there's no way I'm running those tasks from my potato laptop.

Currently, renting an Nvidia Tesla T4, I'm able to generate high quality articles in 6-14 minutes, that rounds up to ~100 articles a day.

Also, I don't know if that's suitable to your script @Sartre, but I'm using this library called PAA 0.0.6, that way, I don't have to pay any SERP api service.

Bon courage pour la suite