Comparison of promotion services by behavioral factors: SERPClick, Userator, Movebo
In Russia (and indeed, around the world, perhaps), there are only three services for artificially "improving" the behavioral factors of sites in the eyes of search engines:
There are many comparisons of these sites on the Internet, but all analyzes come down to a comparison of usability and cost per action.
I wanted to find out which of these services works most efficiently and gives the best result. But the question is how to do this? If you add one site to all services for promotion, it is not clear which of them gave the result. If you add different sites, you will not be able to observe equal conditions for the participants, it’s the same as testing tires from different manufacturers on Lada and Audi, and on the basis of the test to identify the best tires.
Therefore, it was decided to test the services by indirect standards:
1) When artificially improving the behavioral factors for the site, prompts should be wound up in the search
2) In experience, prompts for new queries appear in 3-4 days
3) Prompts are sorted by the frequency of queries recently
Checking the performance and quality of services for wrapping behavioral.
We add “word + unique word” pairs to all services, which are not in the tips and control the appearance of new tips for the given word. The presence of the phrase in the prompts will tell us about the serviceability of the service and the perception of Yandex by its employees.
The word is taken, according to which there are clues, but the frequency according to “! Wordstat ”hints no more than 50. Each service puts the same number of executions on the wrapping of the phrase“ word + service name ”. After a week, the appearance of the tips and the location of the tips are controlled.
For example, let's take the word “seriousness”.
Current hints for this word:
Remove the exact frequency of each request:
Add three phrases to the services, one phrase for each of the services:
sickle click
seriousness user
seriousness moveow seriousness We
will see changes in the prompts in a week and a month later. The presence of a hint and its location will make it possible to conclude on the quality of the service.
For the purity of the experiment, the keyword was not disclosed.
Possible words to use:
Phrases were added to all three services:
promotion without links user;
promotion without links; sickle click promotion without links;
move;
at the time of launch, there were only two phrases in the tips:
promotion without links 2014
promotion without links as
In the user and in the sickle, 100 transitions per month were assigned, but with the move it was not possible to simply assign the number of transitions, because you can set the number of transitions yourself and it’s impossible to manually adjust this value, and the predicted number of transitions is not indicated. Within a week, experimentally, we determined how many domains you need to add to the move, so that 3-4 transitions are made daily. Therefore, by the time when the move became fully operational (08/20/2014), the sickle and user had been working for almost a week.
The prompts appeared:
promotion without links sickplick
promotion without links user
while the sickle click wins, the move did not appear at all, but it was launched later.
Movebo, despite a later launch, rose above the user, and the sickle even surpassed the prompt “promotion without links like”
By chance, an interesting peculiarity was found out - if you make a request from Montenegro, then the prompts and the arrangement of the prompts are somewhat different:
There was no change in the hints from the 27th.
If you make a request from the Netherlands, we will see the same picture as the 27th in Montenegro
Everything remains in place, you can complete the experiment and take stock of
Moscow:
Europe:
1. There were prompts for all three services, which means that all three work with issuance and Yandex perceives them.
2. The best results were shown by Serpklik , the second place after the User , and Movebot on the last.
3. At Serpklik, the average coincidence of the region of the site and the region of the performer is 95%. Apparently, some of the performers were in Europe and this prompted the appearance of a clue there. The user in Europe won - we can assume that he has a lower percentage of coincidence of the site region and the artist. Movebo did not appear at all, well done, but see point 5. Perhaps, an additional experiment should be conducted on the percentage of coincidence of the site region and the region of the performer.
4. All employees of the User are on their own site. Serplik employees on a third-party site (Profittask.com), but the mobo does not disclose where their employees are located, even the technical support on the site did not disclose such information, which suggests that the actions are performed by bots (but they are still taken into account Yandex).
UPD: 09/04/2014, 16:52 Suddenly, the prompts were cleaned up and prompts with the words "sickle click" and "move" disappeared from there. In addition, almost all the tips that were previously in Yandex according to the words “sickle click” and “move” were gone. Hint to moderators
There are many comparisons of these sites on the Internet, but all analyzes come down to a comparison of usability and cost per action.
I wanted to find out which of these services works most efficiently and gives the best result. But the question is how to do this? If you add one site to all services for promotion, it is not clear which of them gave the result. If you add different sites, you will not be able to observe equal conditions for the participants, it’s the same as testing tires from different manufacturers on Lada and Audi, and on the basis of the test to identify the best tires.
Therefore, it was decided to test the services by indirect standards:
Input data:
1) When artificially improving the behavioral factors for the site, prompts should be wound up in the search
2) In experience, prompts for new queries appear in 3-4 days
3) Prompts are sorted by the frequency of queries recently
Testing Tasks:
Checking the performance and quality of services for wrapping behavioral.
The essence of testing:
We add “word + unique word” pairs to all services, which are not in the tips and control the appearance of new tips for the given word. The presence of the phrase in the prompts will tell us about the serviceability of the service and the perception of Yandex by its employees.
Methodology:
The word is taken, according to which there are clues, but the frequency according to “! Wordstat ”hints no more than 50. Each service puts the same number of executions on the wrapping of the phrase“ word + service name ”. After a week, the appearance of the tips and the location of the tips are controlled.
For example, let's take the word “seriousness”.
Current hints for this word:
Remove the exact frequency of each request:
Add three phrases to the services, one phrase for each of the services:
sickle click
seriousness user
seriousness moveow seriousness We
will see changes in the prompts in a week and a month later. The presence of a hint and its location will make it possible to conclude on the quality of the service.
For the purity of the experiment, the keyword was not disclosed.
Possible words to use:
- performance
- functionality
- accuracy
- rationality
- durability
The progress of the experiment:
08/15/2014
Phrases were added to all three services:
promotion without links user;
promotion without links; sickle click promotion without links;
move;
at the time of launch, there were only two phrases in the tips:
promotion without links 2014
promotion without links as
In the user and in the sickle, 100 transitions per month were assigned, but with the move it was not possible to simply assign the number of transitions, because you can set the number of transitions yourself and it’s impossible to manually adjust this value, and the predicted number of transitions is not indicated. Within a week, experimentally, we determined how many domains you need to add to the move, so that 3-4 transitions are made daily. Therefore, by the time when the move became fully operational (08/20/2014), the sickle and user had been working for almost a week.
08/22/2014
The prompts appeared:
promotion without links sickplick
promotion without links user
while the sickle click wins, the move did not appear at all, but it was launched later.
08/27/2014
Movebo, despite a later launch, rose above the user, and the sickle even surpassed the prompt “promotion without links like”
By chance, an interesting peculiarity was found out - if you make a request from Montenegro, then the prompts and the arrangement of the prompts are somewhat different:
09/01/2014
There was no change in the hints from the 27th.
If you make a request from the Netherlands, we will see the same picture as the 27th in Montenegro
09/03/2014
Everything remains in place, you can complete the experiment and take stock of
Moscow:
Europe:
Summary
1. There were prompts for all three services, which means that all three work with issuance and Yandex perceives them.
2. The best results were shown by Serpklik , the second place after the User , and Movebot on the last.
3. At Serpklik, the average coincidence of the region of the site and the region of the performer is 95%. Apparently, some of the performers were in Europe and this prompted the appearance of a clue there. The user in Europe won - we can assume that he has a lower percentage of coincidence of the site region and the artist. Movebo did not appear at all, well done, but see point 5. Perhaps, an additional experiment should be conducted on the percentage of coincidence of the site region and the region of the performer.
4. All employees of the User are on their own site. Serplik employees on a third-party site (Profittask.com), but the mobo does not disclose where their employees are located, even the technical support on the site did not disclose such information, which suggests that the actions are performed by bots (but they are still taken into account Yandex).
UPD: 09/04/2014, 16:52 Suddenly, the prompts were cleaned up and prompts with the words "sickle click" and "move" disappeared from there. In addition, almost all the tips that were previously in Yandex according to the words “sickle click” and “move” were gone. Hint to moderators