Shaunm
Elite Member
- Mar 27, 2014
- 3,437
- 2,998
Morning Guys,
Two days ago this thread popped up on BHW and after reading some of the replies I decided to do a little case study to try and show people what you are really getting when you pay for a "premium" GSA SER List. Previously I had been a big fan of these lists, I have since carried out various private case studies and have came to the conclusion that it is actually cheaper and more efficient to build your own lists.
I have two hopes with this post, first to offer people as much information as possible when they are considering purchasing a premium list and secondly that the list sellers will up there game and create a product worth the price they ask.
For this case study I used Looplines Auto Approve Marketplace, it is important to note that in my experience all list providers have this exact same problem. It is also important to note that some people may have working methods that do not need to be so strict on the platforms they use but where I feel it is necessary I will explain my reasoning behind making various decisions.
Also please note that myself and a few others are currently reporting bugs in SER that cause some platforms that verify not to be added to the verified folder. In this instance the main platform that will effect the end result is Buddypress, in addition to this I am currently unable to carry out the final step of this case study as all public Google proxies I get my hands on are put straight to work on other tasks but I will explain further when we get to that.
View attachment 74236
This is the initial "verified" list synced right from Dropbox, it initially has 339,810 target urls, my first step in processing a list is to remove duplicate URLs as well as duplicate domains. For the benefit of anyone unfamiliar with GSA SER it has a built in feature when removing duplicate domains to ensure that you do not accidentally remove valid blog and image comments that are on the same domain.

This is the list once duplicates have been removed, it now has 334,359 targets. An initial loss of 5,451 (almost 2%) target urls from the list.
Next I move onto removing specific categories. It has long been agreed that RSS, Pingback, Exploit, Indexer and Referrer are a waste of time as they pass very little link juice. In addition as these platforms are easier to post to SER tends to priorities them leaving out the more valuable platforms. Additionally if they are included in your projects then they will clog up SER, your threads, your proxies and your system resources.
I also remove any web 2.0 targets they take up a massive amount of resources within SER for a very very low success rate, if you are wanting to use Web 2.0 platforms in your link building use a dedicated service such as Ranker X.
Since the Google patch to indexing this year (read about it here) that knocked out what was in my opinion the best indexing service around I also now remove Forum, Microblog, Trackback and URL Shorteners. Although these platforms do some link juice my personal testing has shown that it is now in my opinion a waste of time as again they clog up SER, your threads, your proxies and system resources but also now need a fair amount of effort to get them indexed that I am not willing to put in when I am able to use Blog Comments, Image Comments and Guestbooks that are already indexed as my T2/T3 that will push link juice and help with indexing for me.
In addition to these categories I also remove Wikis, my reasoning behind this is that my personal testing has shown both profiles and actual article wikis die extremely quickly when compared to Article Directories and Social Bookmarks so I decided to scrap them from my T1 a while back as I do not want to be wasting resources building out T2 and T3 to a link that will probably be gone within a month or two.
I also remove the Directory category, although many of these are already indexed for you similar to the Blog Comments, Image Comments and Guestbooks there seem to be a lot less auto approve directories when compared to the other three T2/T3 categories that I use. My testing has shown that T2/T3 projects within SER that has directories included in them build significantly less URLs over a set time period when compared to those without directories. If you are just starting out and have a small list then leaving directories in your list should be fine.
Next I remove individual platforms from the remaining categories. My testing has shown that these platforms are mainly no follow so I rather remove them completely to stop them taking up system resources. Even with the remaining platforms and categories you will still be getting between 10-30% no follow rate so in my opinion this does not leave a footprint. The platforms I remove for this reason are.....
ESO Talk
Question 2 Answer
Shownews
Jcomments
Keyword Luv
Blogspot
SPIP
Shoutbox
Kideshoutbox
DRB Guestbook
Guestbook Reloaded
Lazerous
Data so Gallery
Gallery 2
Plogger
Pixlepost
Pixlepost 2
In addition to those platforms I also remove Joomla K2. I know that there is this post over on the GSA forum condemning the use of Joomla K2 and some of my own personal testing suggests that the reasoning behind this is accurate it is not my only reason for not using it. When used around 90% of my T1 was Joomla K2 and I feel that Google may adapt to this soon.
I now remove the General Blog platform at this stage as so many of them are no follow. Again if you are new and your list is small then you an leave them in if you wish and carry out the General Blog step later in the post to make it worth your time.

After pruning the list to my above specifications we now have 35,732 target URLs left. That is a total reduction of almost 90% of the initial list! This is BEFORE The list has even touches GSA SER! I know this list is sold and marketed as a verified list but all that means is that at one point it managed to get through SER into a verified folder. Since that time the web master could have taken any number of steps to either prevent that happening again or at the very lease make it harder for you to put that target through again.
My next step is to load the list into my identified folder in SER. I set up four different types of projects to verify my targets. Two of them target Article and Social Networks (contextuals), two of them target Blog Comments, Image Comments and Guestbook posts (non contextuals). One of each pair is then put through the following filtering system. I make the standard project that loads the target URLs from the identified folder, these projects have the continue to post to previously failed targets box ticked. These two are my main verifying projects and will keep trying to post to failed links on each cycle to get as many as possible. The other is a basic project set up with no ability to pull target links its self. I load this second project type with its target links manually by right clicking the project -> import target URLs -> From Site Lists -> Identified.
The purpose of this second project is to do one run through of the targets and then give the error message no more targets to post to so I know roughly how long it takes for my system to go through the full list. The message appeared after around an hour for the contextual project and three hours for the non contextual project. This means in the 24 hour test period both of the first sets of projects will loop the list a fair number of times bearing in mind the list gets smaller on each loop as it gets more and more links pushed through.
I did not carry out this step for the case study as I find it takes up too much time and resources but if you decided to keep General Blogs in the above filtering then you should now carry out this step. Open the non contextual project that is still running and turn off the general blogs platform. Set up a new project and select only the general blogs platform for it and get it to pull from my verified folder NOT the identified folder. This means that these target links have been pushed through your system at least once so you know they work. Let that project run for a few hours. Next go into your verified folder, open up the general blogs file and delete all the URLs in it, then right click the general blogs project -> show URLs -> Verified.
This loads up all of the verified general blog URLs. Sort them by the do/no follow column. Select all of them that are do follow (normally less than 10%) copy the target URL and paste it into your general blogs file and save it. This means that you now only have the do follow general blogs in my verified folder that have been verified by yourself for your ranking projects to work with. As your list gets bigger you may wish to remove this step as you may find it takes up too much time and resources for the number of URLs you keep.

After the initial run through this is what I have that I consider usable. 1431 verified URLs. that is 0.04% of the original list.

This is what I have after 24 hours of running the projects. 3838 verified URLs. That is around 1.13% of the original list! I honestly think if this was left to run for a week it would probably get to around the 6,000-7,000 mark.As I said at the start, I currently believe there is a bug with SER so out of the platforms I keep mainly Buddypress and Alto CMS none of their links have made it into the verified folder so my guess is there would be around 50-75 additional URLs added to this number at this stage.

Additionally I use an OCR Recaptcha solving service, the above screenshot shows the number of ReCaptcha and Drupal captchas GSA CB has had to send to my OCR service for this test (13,654 total captchas sent to it). If you do not pay the monthly fee for the OCR service then you will get less URLs throughout this process.
On my actual list I also carry out one final step that I'm unable to do for this case study due to using my public Google proxies on other projects and that is take my verified list and run it through Scrapebox to check if the root domain for the contextuals are indexed and the URL for the non contextuals are indexed as if they are not indexed then that site probably has a penalty meaning any links you build on them will never get indexed in Google and may pass a negative effect up your chain.
From my experience of running this final step I find that around 30-40% of the URLs are deindexed and scrapped but as I am unable to do it for this case study I will keep with the presumption that the list will provide around 7,000 total usable links.
I know this is a bit of a long post but if there is anyone still reading then it is time to ask yourself, is it worth spending almost $50 per month for 7,000 usable URLs?
I know people all have their own methods so I am expecting some people to jump in with counter arguments or to try and shoot me down but this is the method I use and I hope it helps a few people.
Two days ago this thread popped up on BHW and after reading some of the replies I decided to do a little case study to try and show people what you are really getting when you pay for a "premium" GSA SER List. Previously I had been a big fan of these lists, I have since carried out various private case studies and have came to the conclusion that it is actually cheaper and more efficient to build your own lists.
I have two hopes with this post, first to offer people as much information as possible when they are considering purchasing a premium list and secondly that the list sellers will up there game and create a product worth the price they ask.
For this case study I used Looplines Auto Approve Marketplace, it is important to note that in my experience all list providers have this exact same problem. It is also important to note that some people may have working methods that do not need to be so strict on the platforms they use but where I feel it is necessary I will explain my reasoning behind making various decisions.
Also please note that myself and a few others are currently reporting bugs in SER that cause some platforms that verify not to be added to the verified folder. In this instance the main platform that will effect the end result is Buddypress, in addition to this I am currently unable to carry out the final step of this case study as all public Google proxies I get my hands on are put straight to work on other tasks but I will explain further when we get to that.
View attachment 74236
This is the initial "verified" list synced right from Dropbox, it initially has 339,810 target urls, my first step in processing a list is to remove duplicate URLs as well as duplicate domains. For the benefit of anyone unfamiliar with GSA SER it has a built in feature when removing duplicate domains to ensure that you do not accidentally remove valid blog and image comments that are on the same domain.

This is the list once duplicates have been removed, it now has 334,359 targets. An initial loss of 5,451 (almost 2%) target urls from the list.
Next I move onto removing specific categories. It has long been agreed that RSS, Pingback, Exploit, Indexer and Referrer are a waste of time as they pass very little link juice. In addition as these platforms are easier to post to SER tends to priorities them leaving out the more valuable platforms. Additionally if they are included in your projects then they will clog up SER, your threads, your proxies and your system resources.
I also remove any web 2.0 targets they take up a massive amount of resources within SER for a very very low success rate, if you are wanting to use Web 2.0 platforms in your link building use a dedicated service such as Ranker X.
Since the Google patch to indexing this year (read about it here) that knocked out what was in my opinion the best indexing service around I also now remove Forum, Microblog, Trackback and URL Shorteners. Although these platforms do some link juice my personal testing has shown that it is now in my opinion a waste of time as again they clog up SER, your threads, your proxies and system resources but also now need a fair amount of effort to get them indexed that I am not willing to put in when I am able to use Blog Comments, Image Comments and Guestbooks that are already indexed as my T2/T3 that will push link juice and help with indexing for me.
In addition to these categories I also remove Wikis, my reasoning behind this is that my personal testing has shown both profiles and actual article wikis die extremely quickly when compared to Article Directories and Social Bookmarks so I decided to scrap them from my T1 a while back as I do not want to be wasting resources building out T2 and T3 to a link that will probably be gone within a month or two.
I also remove the Directory category, although many of these are already indexed for you similar to the Blog Comments, Image Comments and Guestbooks there seem to be a lot less auto approve directories when compared to the other three T2/T3 categories that I use. My testing has shown that T2/T3 projects within SER that has directories included in them build significantly less URLs over a set time period when compared to those without directories. If you are just starting out and have a small list then leaving directories in your list should be fine.
Next I remove individual platforms from the remaining categories. My testing has shown that these platforms are mainly no follow so I rather remove them completely to stop them taking up system resources. Even with the remaining platforms and categories you will still be getting between 10-30% no follow rate so in my opinion this does not leave a footprint. The platforms I remove for this reason are.....
ESO Talk
Question 2 Answer
Shownews
Jcomments
Keyword Luv
Blogspot
SPIP
Shoutbox
Kideshoutbox
DRB Guestbook
Guestbook Reloaded
Lazerous
Data so Gallery
Gallery 2
Plogger
Pixlepost
Pixlepost 2
In addition to those platforms I also remove Joomla K2. I know that there is this post over on the GSA forum condemning the use of Joomla K2 and some of my own personal testing suggests that the reasoning behind this is accurate it is not my only reason for not using it. When used around 90% of my T1 was Joomla K2 and I feel that Google may adapt to this soon.
I now remove the General Blog platform at this stage as so many of them are no follow. Again if you are new and your list is small then you an leave them in if you wish and carry out the General Blog step later in the post to make it worth your time.

After pruning the list to my above specifications we now have 35,732 target URLs left. That is a total reduction of almost 90% of the initial list! This is BEFORE The list has even touches GSA SER! I know this list is sold and marketed as a verified list but all that means is that at one point it managed to get through SER into a verified folder. Since that time the web master could have taken any number of steps to either prevent that happening again or at the very lease make it harder for you to put that target through again.
My next step is to load the list into my identified folder in SER. I set up four different types of projects to verify my targets. Two of them target Article and Social Networks (contextuals), two of them target Blog Comments, Image Comments and Guestbook posts (non contextuals). One of each pair is then put through the following filtering system. I make the standard project that loads the target URLs from the identified folder, these projects have the continue to post to previously failed targets box ticked. These two are my main verifying projects and will keep trying to post to failed links on each cycle to get as many as possible. The other is a basic project set up with no ability to pull target links its self. I load this second project type with its target links manually by right clicking the project -> import target URLs -> From Site Lists -> Identified.
The purpose of this second project is to do one run through of the targets and then give the error message no more targets to post to so I know roughly how long it takes for my system to go through the full list. The message appeared after around an hour for the contextual project and three hours for the non contextual project. This means in the 24 hour test period both of the first sets of projects will loop the list a fair number of times bearing in mind the list gets smaller on each loop as it gets more and more links pushed through.
I did not carry out this step for the case study as I find it takes up too much time and resources but if you decided to keep General Blogs in the above filtering then you should now carry out this step. Open the non contextual project that is still running and turn off the general blogs platform. Set up a new project and select only the general blogs platform for it and get it to pull from my verified folder NOT the identified folder. This means that these target links have been pushed through your system at least once so you know they work. Let that project run for a few hours. Next go into your verified folder, open up the general blogs file and delete all the URLs in it, then right click the general blogs project -> show URLs -> Verified.
This loads up all of the verified general blog URLs. Sort them by the do/no follow column. Select all of them that are do follow (normally less than 10%) copy the target URL and paste it into your general blogs file and save it. This means that you now only have the do follow general blogs in my verified folder that have been verified by yourself for your ranking projects to work with. As your list gets bigger you may wish to remove this step as you may find it takes up too much time and resources for the number of URLs you keep.

After the initial run through this is what I have that I consider usable. 1431 verified URLs. that is 0.04% of the original list.

This is what I have after 24 hours of running the projects. 3838 verified URLs. That is around 1.13% of the original list! I honestly think if this was left to run for a week it would probably get to around the 6,000-7,000 mark.As I said at the start, I currently believe there is a bug with SER so out of the platforms I keep mainly Buddypress and Alto CMS none of their links have made it into the verified folder so my guess is there would be around 50-75 additional URLs added to this number at this stage.

Additionally I use an OCR Recaptcha solving service, the above screenshot shows the number of ReCaptcha and Drupal captchas GSA CB has had to send to my OCR service for this test (13,654 total captchas sent to it). If you do not pay the monthly fee for the OCR service then you will get less URLs throughout this process.
On my actual list I also carry out one final step that I'm unable to do for this case study due to using my public Google proxies on other projects and that is take my verified list and run it through Scrapebox to check if the root domain for the contextuals are indexed and the URL for the non contextuals are indexed as if they are not indexed then that site probably has a penalty meaning any links you build on them will never get indexed in Google and may pass a negative effect up your chain.
From my experience of running this final step I find that around 30-40% of the URLs are deindexed and scrapped but as I am unable to do it for this case study I will keep with the presumption that the list will provide around 7,000 total usable links.
I know this is a bit of a long post but if there is anyone still reading then it is time to ask yourself, is it worth spending almost $50 per month for 7,000 usable URLs?
I know people all have their own methods so I am expecting some people to jump in with counter arguments or to try and shoot me down but this is the method I use and I hope it helps a few people.