Case Study Determining Which Parasite is Easiest to Rank

Discussion in 'Case Studies and Journals' started by cardine, Dec 18, 2015.

Share This Page

  1. cardine

    cardine Administrator Staff Member

    Joined:
    Dec 9, 2015
    Messages:
    1,064
    Likes Received:
    1,026
    I've been talking a lot recently, both publicly and privately, about the effectiveness of ranking parasites, and how generally easy it is to do. I often get people asking me what parasites they should be picking, and it got me thinking if we could test on our own to see which parasites Google inherently trusts the most.

    A lot of people just look at the parasites that show up the most in Google, but this can sometimes tell an incomplete picture. There might be certain parasites that Google trusts a lot because they are underused.

    So I am going to test this for myself to see what parasites Google likes the most and what parasites Google likes the least. To do this we are going to be doing several different things:

    The first thing I am going to be doing is I am going to be identifying all of the parasites I want to check. The criteria I am using for parasites are as follows:
    1. There is no moderation for new accounts or content.
    2. There is a way to fairly easily put an affiliate link or ad on the page.
    3. This is a big site that likely has some trust with Google.

    Following this basic guideline I have compiled a basic list of parasites:
    Since this list might be incomplete, over this weekend I will be running an analysis across over 2 million different SERPs to see what the most commonly ranking domains are. I'll then be applying the above criteria to any other common domain names that script finds to see if there are other parasites I missed. Additionally I'll keep an eye out here if anyone reading this has any suggestions for parasites that I might have missed.

    Once that list is compiled, the real experiment will start. Each parasite will be built out and ranked for the same dummy keyword, and we will be able to follow the progress for each keyword live. In order to make the process as scientific as possible I am going to try to make sure there are as few external variables as possible. So I will be placing the following controls on the experiment:
    1. I will be picking a dummy word with virtually zero competition so that we will be able to see only these parasites competing with each other. Seeing how well these parasites do compared to other established sites might be the subject of a different case study, but I want to see exclusively how each parasite does compared to each other for this experiment.
    2. Since it will be a dummy keyword, each account will have the username or name be just the dummy keyword so that no page has an unfair advantage because of relevance (for instance if the keyword was "dog snuggies" some properties might not have that username available).
    3. I will write out a dummy article and a dummy description and heavily spin it. Each parasite that would require an article will get the dummy article and each parasite that will require a description will get a description. For this example I am going to do spun content all from the same seed article - I don't want one parasite to rank better because one seed article was better liked by Google than another. I was considering just having each one have the same duplicate content, but I'm worried that will lead to Google handling this a little differently. If people disagree on this let me know!
    4. At first I will send no links at all and we will follow the progress. Then after the rankings have stabilized I will report the results and then a small number of links will be built to each page. My hope is that because of the lack of competitiveness, we will see a lot of sites in the top 10/20 without any links which should give us a good idea about which parasites Google trusts the most. From there we can then see how Google responds when each one has links pointed to it.

    At the end of this we will get to see how each parasite performs and ranks. From there we should have a good list of what parasites Google likes the best when all other factors are the same. We can then also compare which parasites did that best in the case study versus which parasites show up the most in Google, and from there we can find which parasites are the most untapped (meaning they rank really well but people rarely try to rank them).

    To make sure that there is no interference I won't be sharing the keyword during the case study, but I will be posting rank tracking graphs so we can all track together which parasite is performing the best.
     
  2. Golan

    Golan Established Member

    Joined:
    Dec 10, 2015
    Messages:
    101
    Likes Received:
    84
    This will be very cool case study. But - how do you find time for everything? ))
     
    Dipesh likes this.
  3. dzianis

    dzianis Moderator Staff Member

    Joined:
    Dec 15, 2015
    Messages:
    230
    Likes Received:
    370
    Great case study idea!

    I see a potential problem at this stage - some parasite properties require an extra push to get indexed. I see some like that in your list and there will be such sites especially in that scrape you are running. Some hosts give you orphan pages, some an orphan subdomain - so when you run the first ranking check you may see half of these ranking and the other half not even indexed.

    You could do something to all of them to get bots crawling right after you create them, e.g. tweet with some trendy hashtag - just to give them equal chances from the start.
     
  4. cardine

    cardine Administrator Staff Member

    Joined:
    Dec 9, 2015
    Messages:
    1,064
    Likes Received:
    1,026
    That is a good point. My original plan was to do no link building and see and then do some link building (which would inevitably index all of the other sites) but I agree that it is a silly condition to do no indexing at all since that would never really happen in real life. I'll make sure to do something (the same thing) to each site to get them indexed. Whatever it is (I think a tweet with a trending hashtag is a good idea), I'll make sure to document it here.
    My secret!

    In all seriousness, I am pretty good about keeping myself on track, and honestly a case study like this will feel like a break from all of the other work I do. Also @megodon will be helping with some of the nitty gritty details, like creating each property.
     
    Dipesh and Jcow or Jwow? like this.
  5. cardine

    cardine Administrator Staff Member

    Joined:
    Dec 9, 2015
    Messages:
    1,064
    Likes Received:
    1,026
    Update:

    We've been running our analysis on most popular parasites, and we have now compiled our automatically generated list of the most common parasites. This list was taken from the most common sites that show up across over 2 million SERPs that allow for (fairly easy) user submitted content:
    So that list by itself should be very useful for uncovering good sites that can be used as parasites. Some might be slightly niche specific, but all of them have proven ranking power.

    However for this case study we want sites that are very easy to post to, so something like foursquare.com which requires a valid location, or glassdoor.com which requires a valid company might be a little outside of the scope of what we can realistically do. So by combining our original (manually created by me) list and the "easy to create" properties from our automatically generated list we have the following "pretty easy to create" parasites:

    Now that we have the full parasite list we are going to be preparing content for each of these properties. Then starting next week we will create each parasite as quickly as possible. My goal is that every single one of the above properties is created in under a day.
     
    Anaconda, dtea, emp and 4 others like this.
  6. babuni

    babuni Member

    Joined:
    Dec 17, 2015
    Messages:
    14
    Likes Received:
    8
    @cardine, would you discriminate dofollow and nofollow of such parasite a.k.a social network and social media sites.
    I have found most of the web 2.0 properties works nicely for Google. What we face is the indexing of content on those sites.
     
    Last edited: Dec 28, 2015
    Dipesh likes this.
  7. Bizzy

    Bizzy New Member

    Joined:
    Dec 24, 2015
    Messages:
    9
    Likes Received:
    5
    @cardine , I wouldn't use weebly unless you go for one of the paid plans, as the free plan blocks all bots. Your weebly parasite won't get indexed correctly.
     
  8. dzianis

    dzianis Moderator Staff Member

    Joined:
    Dec 15, 2015
    Messages:
    230
    Likes Received:
    370
    It does not matter whether a parasite gives a do- or a no-follow link, the point is to redirect visitors to the money page, so any link is a link.

    This statement feels wrong.

    Let's see if there are any weeblies ranking:

    [​IMG]

    Just to check, I took a weebly site of my own that I know is made with a free account:

    [​IMG]


    Yep, your statement was wrong.
     
    John Stover, dtea, sites and 2 others like this.
  9. cardine

    cardine Administrator Staff Member

    Joined:
    Dec 9, 2015
    Messages:
    1,064
    Likes Received:
    1,026
    The point of this case study is not to see which parasite has the most link juice. Rather, the point of this case study is to see which parasite can rank the easiest on its own.

    So we aren't doing this to figure out which parasites you should use for link building, but rather which parasites you should be trying to rank. This is great for blackhat SEO tactics where you take advantage of a parasite's built in trust, as well as for reputation management.

    So, as @dzianis said, it shouldn't matter whether a link is a do follow or not - the parasite is acting as a landing page, not a linking property.

    For this case study each property will be created manually. Since we aren't building these as link properties it shouldn't be an issue to go in and manually create each one.
     
  10. Bizzy

    Bizzy New Member

    Joined:
    Dec 24, 2015
    Messages:
    9
    Likes Received:
    5
    @dzianis I'm not talking about old pre-existing weebly properties. I've got loads of them & they are indeed indexed.

    Try registering a brand new property with a brand new username, and the free plan won't get your site indexed.
    Don't believe me ? Take a look at the weebly pricing page.

    weebly.JPG
     
    sites and cardine like this.
  11. cardine

    cardine Administrator Staff Member

    Joined:
    Dec 9, 2015
    Messages:
    1,064
    Likes Received:
    1,026
    Have you created a Weebly since they made that change where you can confirm that is the case? Just want to make sure "Search Engine Indexing" doesn't refer to them indexing it for you as opposed to the free plan actively preventing indexing.
     
  12. Bizzy

    Bizzy New Member

    Joined:
    Dec 24, 2015
    Messages:
    9
    Likes Received:
    5
    I had a few created recently. Here's one created a month ago, on the 28th of november :
    weebly-goog.png
     
    sites and cardine like this.
  13. dzianis

    dzianis Moderator Staff Member

    Joined:
    Dec 15, 2015
    Messages:
    230
    Likes Received:
    370
    I checked a couple that were made in December. Yeah, robots.txt is disallowing it from all bots BUT they are getting indexed nonetheless, just like on your screenshot above. It does not matter that much what the robots file says since Google still crawls them, indexes titles and reads all the content, and counts the links. For the parasite test purpose it matters even less because visitors can click through to the site from Google SERPs.
     
    sites, cardine and Golan like this.
  14. youknowwho

    youknowwho Member

    Joined:
    Dec 25, 2015
    Messages:
    15
    Likes Received:
    10
    You can do all this or monitor your niches on SERPwoo and find parasites which rank .

    Edit : Here is the SERPwoo post on how to do it
     
  15. cardine

    cardine Administrator Staff Member

    Joined:
    Dec 9, 2015
    Messages:
    1,064
    Likes Received:
    1,026
    I already posted a list of which parasites rank the most. However which parasites rank the most is not the same as which parasite is easier to rank. An underused parasite might be far easier to rank because it is underused - Google might trust it more as a result.

    Or to put it another way, a list of the top ranking parasites will show you correlation but it will not show you causation. This case study is going to try to find causation.
     
    sites, mstchr, Insaint and 2 others like this.
  16. dzianis

    dzianis Moderator Staff Member

    Joined:
    Dec 15, 2015
    Messages:
    230
    Likes Received:
    370
    Well, a truly underused parasite won't make it to your initial afterscrape list, because it is not repeating in serps that often :) At least I'm kinda happy to see you didn't mention ones I'm actively ranking lately :)
     
  17. cardine

    cardine Administrator Staff Member

    Joined:
    Dec 9, 2015
    Messages:
    1,064
    Likes Received:
    1,026
    That is true! The cutoff was largely decided based on how much manual time it would take to create each of these properties - I wanted to be able to create them all in under a day off of one spun article which added some constraints. Depending on the results of this case study we might have to revisit it again with a broader list of parasite properties.
     
  18. cardine

    cardine Administrator Staff Member

    Joined:
    Dec 9, 2015
    Messages:
    1,064
    Likes Received:
    1,026
    Update:

    We have (well... mostly @megodon) created all of the properties... we ended up having to get rid of a couple of sites from the list that didn't allow for easy account creation, so here is our master list:
    We have not done anything to index any of these yet, so tomorrow we will look at which ones got indexed and then we will index all of them.
     
  19. cardine

    cardine Administrator Staff Member

    Joined:
    Dec 9, 2015
    Messages:
    1,064
    Likes Received:
    1,026
    Update:

    We now have our first day of rankings!

    upload_2015-12-30_11-13-8.png

    These rankings all occurred before we did any indexing. So this is without any link building or any indexing whatsoever, so many of these pages are not ranking because Google likely still doesn't even know they exist.

    We just indexed each URL, so the rankings we get over the next couple of days should start to become more complete. Still no links have been built to any of these pages, so these rankings will be based entirely off of any automatic interlinking done by the parasite and Google's own trust for each parasite.
     
    darnoldi, sillyseo, yns01 and 5 others like this.
  20. sillyseo

    sillyseo New Member

    Joined:
    Jan 2, 2016
    Messages:
    1
    Likes Received:
    0
    any update?
     
    Last edited by a moderator: Jan 2, 2016