BULK Feature for creating articles relating to one topic but stuffed with special keywords in it?

henningnet

Hey guys,
hey Sven,

lets say my overall topic is "Online Casino" (which is perfectly scrapable, because there is tons of content on the internet).

Now I have 5 longtail keywords, some of them might be completely unused, nobody ever wrote about them, so also nothing might be scrapable for them.

Lets say the keywords are those:
1. best blue casino bonus
2. best red slot bonus
3. best green online slot
4. best blue casino design
5. best online casino webdesign

So my idea was that GSA could scrape 5 articles in general about "Online Casino" but stuff each of those 5 keywords into one of the articles + sub-headings + title.

So that I get 5 different articles that are in general about Online Casino, but the first one includes "best blue casino bonus" several times, the second one includes "best red slot bonus" several times and so on.

My idea behind this is:
Often I tried to scrape longtail keywords for certain topics. But GSA was not able to create articles for that keyword, as there is not enough content about that very special topic or even nobody wrote about it. So my idea was to scrape content from more "general topic", in this case its "Online Casino" and then just let GSA stuff the needed keyword several times into the article, as this should do it for Google and BING to rank it.

So not sure if this is already possible or if this can be added as a feature?

Thanks a lot, as always!

Sven

Thats perfectly possible.

The keyword "Online Casino" is defined as a keyword to scrape only (IB column). Then you add the rest of the keywords with at least "OB column" being checked + you use options to add keywords + use decoration with keywords.

Minidou

Hey,

Sven said:

Thats perfectly possible.

The keyword "Online Casino" is defined as a keyword to scrape only (IB column). Then you add the rest of the keywords with at least "OB column" being checked + you use options to add keywords + use decoration with keywords.

By doing so, will the resulting article have the keyword "best red slot bonus" (to take one of the example mentionned above) in the Title + inside some Hx + inside the content ?

From what I understand of the way GSA CG is currently working, it will not be the case, and all what we'll end up with will be an article with "best red slot bonus" inside the content, but neither in the Hx nor the Title.

If I understood henningnet right, the idea suggested here would be to create an article using a keyword which have insufficient ressources to be properly scraped, and artificially stuff it in the Title + Hx + Content in order to push said keyword.

Currently however, it's not possible to insert a keyword in the Title, because the tool create its Title from existing articles scraped from the SERP. It can't create Titles from scratch (an idea I also proposed in the past).
In this case, since there are no "best red slot bonus" articles in the SERP, we would never end up having Titles containing "best red slot bonus".

As for the Hx, if I'm not mistaking, the tool will pick up random keywords to build the differents sub-headings, but not necessarily the keyword we're looking to push ?
If we have, say 1000 keywords (with the OB checkbox checked), with "best red slot bonus" among them, we could totally end up with an article having Hx such as "best gambling service", "online jackpot", "earn 1000€ with online casino", but never "best red slot bonus".
Is that right ?
On this one I'm a bit lost because there are many options in the "Insert random paragraphs spoilers/decorations" option, which have an impact on how the Hx are designed in the end-result, but these options are not clearly explained. I don't know how they work exactly... That's something that would need to be explored further in the FAQ, when you'll have time.

The main problem here is that the generation of the Title/Hx/keywords in content are not linked together, these functions work independantly. Hence, we end up with heterogeneous content, which is not optimal in order to rank on rare longtails keywords.

The goal would be to end up with a single keyword, taken from all of our list, being focused on when building an article : The tool would need to take "best red slot bonus" to build the Title, build some sub-heading (not all of them, of course, but at least one H2/H3, and with variations in order not to do keyword stuffing, for example by combining "best red slot bonus" with another keyword), and insert it in the content.


P.S. : I think the way the OT column is working would however come in conflict with this feature.
"Keyword should be used in Resulting Title" implies that the keyword whose OT checkbox is checked will end up in the Title, but if you ever implement our idea, that would mean you'd have to rework how the tool work to build the Title (because in the example mentionned above, it would be impossible to build a "best red slot bonus" Title).

In fact, I come back to my initial proposal to redesign the way titles are created, by proposing two new options among the existing columns: One to be checked indicating our wish to reuse the SERP titles as they are, and another indicating our wish to create titles "from scratch".

When these two options would be checked, the first option (create titles from those already existing in the SERP) would have priority over the second one).

This should of course be explained in an infobox and in the GSA CG FAQ that you are writing (congratulations for the work by the way!).

Sven

you are wrong...you can get titles and keywords in the body with the new column options per keyword as all explained here: https://docu.gsa-online.de/content_generator/project_settings#input_-_keywords

draculax

hello,

I confirm we can have the keywords without the title now thank you sven.

On the other hand how to have this keyword in h2 and h3 .

I do not understand how to generate h2 and h3 it is a mystery to me.

it would be necessary that one could choose the number of h2 or h3 inserted in the following text the number of paragraph and words .

Bravo for the notice I just discovered it is really great.

I'm waiting for the black friday to make a new purchase from you.

Sven

@draculax you need to enable "Insert random paragraph spoilers/decorations"

draculax

hello,

yes I do like that but there is a lot of randomness.

How to know if an article will have an h2 and an h3?

How to insert the keywords in the h2 and h3 .

I don't understand how it works 

thanks you sven 

Sven

Well you would uncheck all on that dialog but "Use <h2> TAG" + "Use <h3> TAG" + "Take longtail keywords" + "Take generated titles from keywords".

draculax

ok thanks sven .

I'll try right away. 

thank you

Minidou

Hello there,

I've tested a few things and it appears there's no way for me to make it so that my content have ONE keyword focused on to build my articles (I mean, Title + Hx + Content).
The tool picks up several keywords to build my articles, but doesn't focus on a single one (which would consolidate the semantics and help better rank on the focused keyword).
This is especially crucial when the tool tries to build an article using a keyword with few entries in the SERPs.

I might have misused the options so far ?


All I'm getting so far is the following :
1-The tool take a keyword, create a Title with that keyword (ONLY if it has found an article in the SERP having that keyword in its Title AND if it has scraped enough data to link it with the keyword in question, which is limiting when it comes to rares keywords).
2- Build a first sub-heading (H1 or H2) with the same sentence built for the Title (with no variations whatsoever, which is not optimal imho)
3- Build the rest of the article.

Sometimes, there are strictly no references to the keyword elsewhere : The keyword is nowhere to be found in the content or in the subsequent Hx.

Example : I have these keywords defined as scrap keyword : "marketing", "email marketing", "copywriting and sales".
As OT/OB keywords, I have "copywriting", "emailing" and "What advertising tells us about the world".
What happens systematically is that "What advertising tells us about the world" will NEVER appear in the generation, probably due to the fact that the tool did not find enough cross-referenced data, unlike "emailing" and "copywriting" which are dominant.

Because of this, generally speaking, lots of longtails keywords will simply never be taken into account when building the articles (because there are not enough crossed data).

In order to prevent that from happening, what was proposed earlier was to redesign a bit how the tool build the articles when not enough datas are found, by artificially inserting the rare longtails keywords in the article.

But in order to do that, of course, we should have to specify that it's our wish to do so, with the help of a new option for example ("Force Inserting Keywords" ?).
Indeed, some users might not want to demolish the readability of their article ! It's important to let them choose how the tool will build their article.
The option described above to artificially insert rare longtails keyword is only a key feature when it comes to spamdexing, where readability for visitors is not a relevant consideration.

draculax

I agree with you mimidou.

we should have this kind of option to improve the semantics especially in markov.

ElHermano

yeah minido have good ideas but dont forget there are alot of gsa product and only one dev (i think ?), and most of the ideas probably require lot of work because if i understand minidou's ideas right
lot of things would need to be reworked, so i dont think these ideas could be implemented sadly 
but good work to you sven anyway, you have done alot in the tool, all alone!

Sven

Even though Im not replying instantly for these suggestions, I keep them open on my to-do list and have already some ideas on how to get around the keyword on topic vs on body issue...I hope I can get to it next week.

But yes, Im the only developer for almost any GSA products. I do my best to keep you all happy. But right now I have to concentrate on GSA Keyword Research (the new baby) beside fixing bugs or adding small improvements on other tools.

draculax

good courage and well done you are a warrior Sven

Minidou

Good luck Sven ! I am fully aware that the development is difficult to manage on your own, I've never expected my suggestions to be put into shape immediately of course !

These are things that require time, and it's perfectly understandable to focus on the core projects.

I just share a few ideas on the fly before I forget them, as I use the tool

henningnet

Hey guys, so me as a OP I am sorry that I did not reply the last days. But I read through all the comments and for me it looks like that Minidou and I search for a similar feature.

Sorry upfront if I might repeat myself here...

So if I could give GSA a "big" topic like "Online Casino" plus a list of lets say 100 Longtails. And the tool would scrape me 100 different articles about "Online Casino" and just stuff 1 of the 100 keywords into each of the articles content, h2, h3 whatever, that would be nice.

So I would end with 100 pieces of content about Online Casino and each of those 100 articles "concentrates" on one focus keyword out of the list of longtails I provided. As those longtails might not be scrapable itself (because not enough data available to scrape). So that I have one article per longtail

I also tried it with the actual version of the tool and I ended with content that mixes all the longtails and there is really no 100% concentration on 1 longtail, at least when you use it bulk.

I hope we are all on the same topic and I would really love to get this feature, because I often end with "no content" when it comes to tricky longtails that nobody has written a piece of content about.

Again, what Sven creates as a software is pretty awesome, I am not sure how much I recommended one of the tools (besides the classic Search Engine Ranker) to friends in the business.

Thanks!

Sven

I am working on that. I really thought I get to this during the week but had to deal with other stuff. I will get to it next week I hope.

draculax

ok thanks you Sven 

henningnet

Sven said:

I am working on that. I really thought I get to this during the week but had to deal with other stuff. I will get to it next week I hope.

Superb! Thank you!

henningnet

Hey guys,
hey Sven,

just wanted to ask about an update for this.

Thanks a lot

Sven

sorry, needs to wait. I have this thread opened here in browser and will get through it ones I promise!

henningnet

Hey Sven,

thank you for updating this thread. I appreciate that.

Looking forward

Sven

The latest update should improve the content quality a lot.

The created content is much more based on the keyword chosen in article title. I can however improve it a bit more when "Mix Paragraphs from all articles" is used as algorithm, but it should create much better content now as well and especially for all other methods.

FrancoisK

Hi Sven,
This is very good news
As I talked to you in another post, it's to know if it's possible to have an option not to mix several keywords in a post for a good density of words in the generated article.
Francois

henningnet

Thanks a lot. I will try it like Monday or Tuesday and let you guys know how it worked

jo00895

Hi  guys,

Have you performed any tests since the last update ?

When I do an articles generation (with the parameters suggested by Sven) overall it works fine but in the result of my articles I see that there is only one source that was used.

Sven

what algorithm do you use?

jo00895

Mix Paragraphs from all articles

Sven

can you provide the project backup for review? I guess the paragraphs from that article are just matching best in favor for anything else.