In this post, I will be testing how efficient list filtering is when used on projects building contextual links.
So Whats The Plan?
As with my non-contextual link case study I plan to have an instance of GSA Search Engine Ranker dedicated to the test running nothing but 25 test projects to ensure 100% of system resources available for the test are put into the tests. The test will be conducted on the exact same VPS I used for the non-contextual case study so readers are able to see the different in link output between contextual and non-contextual engines with the same hardware. As these engines are contextual and will require link building to them for live projects I plan to report on both the links per minute of the test as well as the verified per minute of the test.
One thing I will be changing is the number of tests I will be completing. Previously I completed three tests and I feel this is fully sufficient for non-contextual link engines, this time, due to testing contextual link engines I am increasing it to five tests.
Although I usually only let my live projects pull from my do follow only folder that is just my personal preference so for these tests I will be letting the projects pull from both my do follow and no follow folders. Similar to the non-contextual test, the filtered versions of the list have had their filtering projects running for around 12 hours to try and keep the comparison between contextual and non-contextual link output fair.
The first test will be the same as my non-contextual case study based around an completely untouched premium target list exactly as it is delivered to users from the list provider.
The second test will be the same as the non-contextual test too featuring a list filtered to platform level. This means that I will remove all of the engine files that do not belong to either the articles, social networks or wiki platforms.
The third test is the same as the third test from the non-contextual test featuring a list that has been fully filtered using the filtering process I explain in this post.
The fourth test is a new test as I will be using a fully filtered list similar to the third test but I will be removing the Joomla K2, PHPWeb and Media Wiki engine files for the run as I do not use them in live link building for various reasons.
The fifth and final test will essentially use the exact same engine targets as the fourth test meaning a fully filtered list with the Joomla K2, PHPWeb and Media Wiki engine files removed but rather than letting the project batches automatically verify links as and when they wish to, link verification will be turned off completely.
This is exactly how I run my live link building projects, the reason for doing this is to optimize the total daily link output of GSA Search Engine Ranker. Turning automatic link verification off for the project enables all of the projects to focus on building as many links as humanly possible without randomly going into verification most. Sometimes my live link building projects will be set to verify after a custom time set in the project options tab but I usually turn it off completely and manually do it when I need it doing.
Due to link verification being turned off this fifth and final test will only be reporting on the links per minute metric in an attempt to see if my theory is correct and turning verification off does increase the potential link output of the tool.
Again, I plan to let each of the tests run for one hour before recording their data, rebuilding the test projects and moving onto the next test.
Although I predict that the project with automatic verification turned off will display the highest links per minute and have a higher total link output over the course of a day than the other projects I will be leaving it out of my prediction as I have no way to know the verified per minute it is actually producing to compare it to the other test batches.
With that being said, I predicted the test groups will rank as below shown in predicted decreasing order by verified links per minute.
- The Fully Filtered List With Joomla K2, PHP Web and Media Wiki Included.
- The Fully Filtered List With Joomla K2, PHP Web and Media Wiki Removed.
- The Platform Filtered List.
- The Untouched List.
I predict that the list with the Joomla K2, PHPWeb and Media Wiki engines left included will beat the one without due to the fact it has more valid targets to post to. The PHPWeb Engine also requires no email verification and thus removes a step from the process most other contextual engines require meaning GSA Search Engine Ranker is able to complete these targets quicker, they are usually an instant verification too similar to the Media Wiki engine.
The Joomla K2 engine seems to have had some things changed on its content management system recently, previously it operated in a similar way to PHPWeb where it would be an instant verification but it now holds the links in its submitted tab for some time before the become verified, although I am not sure if many of the Joomla K2 targets will be verified in one hour due to this I am confident that the PHPWeb and Media Wiki engines will compensate for this resulting in a higher link output.
The Results Are In!
The Untouched List
As you can see from the screenshots below, the untouched list came in at 136.52 links per minute, 94.59 verifieds per minute and a total of 5624 verified URLs after an hour.
The Platform Filtered List
As you can see from the screenshots below, the test batch filtered to platform level came in at 146.56 links per minute, 95.35 verified links per minute and 4963 total verified URLs after an hours run time.
The Filtered List Including Joomla K2, PHPWeb and Media Wiki
As you can see from the screenshots below, this test came in at 344.49 links per minute, 297.23 verified links per minute with a total of 18512 verified links build in the allocated hour.
The Filtered List With Joomla K2, PHPWeb And Media Wiki Removed
As you can see from the screenshots below the test for the projects running a filtered list with the Joomla K2, PHPWeb and Media Wiki engines removed resulted in 381.01 links per minute, 225.85 verified links per minute, with 13785 total verified URLs after an hour.
One thing I would like to point out in the below screenshots is the number of projects currently in verification mode in the screenshots. For those new to GSA Search Engine Ranker this is displayed with the blue background behind the status of the project. I hope this helps you understand how much potential link output you are losing by having your projects be able to swap to verification mode as they please rather than having you manually tell them when to do it.
Out of the 22 visible projects, 12 are currently in verification mode, thats 55% of the visible projects taking a break from producing links…
The Filtered List With Joomla K2, PHPWeb And Media Wiki Removed And Verification Turned Off
As you can see in the screenshots below, the test with the Joomla K2, PHPWeb and Media Wiki engines removed as well as having automatic re-verification turned off came in at 516.36 links per minute, 305.07 verifieds per minute and 18966 verified links created in an hour.
Although I expected the links per minute of these test projects to be higher than the previous test batch I did not expect the verified links per minute to also be higher although I did expect a verified links per minute count.
This is because some of the engines have automatic verification, this means as soon as you submit a link it is verified and GSA Search Engine Ranker detects this verification and adds it to your verified link count.
My theory is that because the overall links per minute has increased by forcing the projects to not go into verification mode, the projects were able to submit more links to the engines that automatically verify. As you can see the majority of the verified links in the last 1000 verified links pane to the right of the screenshots is showing a majority of XpressEngine.
This is due to automatic verification being turned off, do not think that the link profile will be made up of 99% XpressEngine. If these projects were building links to my money site network or my test network then when I manually set these projects to verify their links at the end of the day I would expect the last 1000 verified links to be filled with engines such as WordPress, Drupal, BuddyPress and PHPFox as these ending require the link to be actually verified as well as a bunch of additional engines balancing the link profile of the projects out.
Simplifying The Result Data
Organized By Links Per Minute In Descending Order.
Organized By Verified Links Per Minute In Descending Order.
Organized By Total Verifieds In Descending Order.
What I Make Of It All
As you can see, the test batch without automatic verification came out way ahead of the bunch proving my theory that this is the best way to run contextual projects. Even though automatic verification was actually turned off for the test batch it still came out on top for verified links per minute and total verifies built in the hour due to the ending with automatic verification.
Here is a comparison when comparing the best performing test group against the worst performing for each metric breakdown.
The group with automatic verification turned off had an increase of 278.23% in links per minute when compared to the group using the unfiltered list.
The group with automatic verification turned off had a minimum increase of 222.51% in verified links per minute when compared to the group using the unfiltered list. Due to automatic verification being turned off the actual verified links per minute will actually be higher than this in reality.
The group with automatic verification turned off had a minimum increased of 282.14% in verified links created in the hour test period when compared to the list filtered to platform level. Again, due to automatic link verification being turned off the total increase will be higher than the recorded numbers.