1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Kicking your ass with data.

Discussion in 'White Hat SEO' started by validseo, Feb 22, 2017.

  1. validseo

    validseo Jr. VIP Jr. VIP Premium Member

    Joined:
    Jul 17, 2013
    Messages:
    917
    Likes Received:
    527
    Occupation:
    Professional SEO
    Location:
    Seattle, Wa
    Before you complain about sample sizes... understand I measured and calculated over 50,000 data points to produce these charts by measuring over 300 factors from 100 web pages for the search term "SEO software" (So you can check my work and reproduce my results if you want to). The important part is that the correlations are statistically significant which I have demonstrated in the last image at the bottom. Obviously you need to redo the analysis for different search terms, but I anecdotally see these correlations for many different terms.

    Having the year in your page title appears to pack a punch.

    upload_2017-2-22_10-23-59.png

    But having the year in your URL does not. Hmmmm...

    upload_2017-2-22_10-24-5.png

    Despite Moz claiming that HTTPS is growing in usage in top results it still doesn't appear to correlate with rank position. As many have suspected HTTPS bonus appears to be dead.


    upload_2017-2-22_10-24-12.png

    Who says Google doesn't play favorites? :)

    upload_2017-2-22_10-24-20.png

    Matches in H1 tags don't correlate and matches in general in H2 tags don't correlate BUT leading matches in H2 tags pack a punch! :)

    upload_2017-2-22_10-24-26.png

    Just showing my work for the data police! :)

    upload_2017-2-22_10-25-26.png
     
    • Thanks Thanks x 25
    Last edited: Feb 22, 2017
  2. BloodyNinja

    BloodyNinja Power Member

    Joined:
    Oct 28, 2013
    Messages:
    605
    Likes Received:
    586
    Location:
    Deeptown
    Great work! Data rules the world

    What are "leading matches in h2" ?
     
  3. validseo

    validseo Jr. VIP Jr. VIP Premium Member

    Joined:
    Jul 17, 2013
    Messages:
    917
    Likes Received:
    527
    Occupation:
    Professional SEO
    Location:
    Seattle, Wa
    Anytime the content of the H2 tag begins with a match.

    Leading Matches in H2 tags:
    <h2>SEO software</h2>
    <h2>SEO software is blah blah blah</h2>

    Keep in mind the Google considers the following to be matches:

    seo
    seo software
    seosoftware
    software

    Google tells us this by making matches bold in the search results:
    upload_2017-2-22_10-52-34.png

    An even better example of this is when you search Google for DUI lawyer los angeles upload_2017-2-22_10-55-24.png
     
    • Thanks Thanks x 2
  4. BloodyNinja

    BloodyNinja Power Member

    Joined:
    Oct 28, 2013
    Messages:
    605
    Likes Received:
    586
    Location:
    Deeptown
    Thanks! This is a really useful insight and makes sense from the common sense point of view as well.

    I really wonder what would happen if one abandoned the usage of h1 completely. In some CMSs, h1 is used for a meaningless thing like site-wide title while article titles are encapsulated in H2s. Would they have an 'unfair advantage' over websites with standard composition or not...
    Will try playing around with titles on my websites.
     
  5. validseo

    validseo Jr. VIP Jr. VIP Premium Member

    Joined:
    Jul 17, 2013
    Messages:
    917
    Likes Received:
    527
    Occupation:
    Professional SEO
    Location:
    Seattle, Wa
    H1 is very important in Bing. I don't recommend abandoning it.
     
    • Thanks Thanks x 1
  6. ThopHayt

    ThopHayt Jr. VIP Jr. VIP Premium Member

    Joined:
    Jul 25, 2011
    Messages:
    6,714
    Likes Received:
    2,557
    Very nice findings, thank you for sharing these with us.

    Did you find any correlations having to do with page DESCRIPTIONS by any chance??? Length, keyword contents, etc?

    -ThopHayt
     
  7. Hawkster

    Hawkster Jr. VIP Jr. VIP

    Joined:
    Jun 22, 2013
    Messages:
    3,514
    Likes Received:
    3,730
    Gender:
    Male
    Occupation:
    Listen to everyone - Follow no-one
    Location:
    UK
    Home Page:
  8. validseo

    validseo Jr. VIP Jr. VIP Premium Member

    Joined:
    Jul 17, 2013
    Messages:
    917
    Likes Received:
    527
    Occupation:
    Professional SEO
    Location:
    Seattle, Wa
    I do measure those... no strong correlations. "Result Summary" refers to the description snippet that appears in the SERPs versus "Meta Description" which is the content put in the tag on page.

    Their correlation coefficients range from 0 to -0.28 which favors the null hypothesis that they are not compelling ranking factors. They are critical for CTR though so don't mistake that for being unimportant. And I do see them correlate from time to time for other keywords so they may just be very weak factors which would require more samples to properly detect.

    But weak factors aren't useful to me... I want things that have the best chance to move the needle. :)

    CORRECTION: While they didn't have strong correlations over all top 100 results the Result Summary matches did correlate by average measure per page at -0.73 which IS significant... So match count in the description snippet in the SERPs does correlate with rank position.

    upload_2017-2-22_13-50-30.png


    upload_2017-2-22_13-50-42.png

    upload_2017-2-22_13-50-52.png

    upload_2017-2-22_13-51-2.png
     
    • Thanks Thanks x 1
    Last edited: Feb 22, 2017
  9. sachin_panwar1996

    sachin_panwar1996 Newbie

    Joined:
    Nov 7, 2016
    Messages:
    8
    Likes Received:
    1
    Gender:
    Male
    which serp analysis tool is it
     
  10. Nargil

    Nargil Jr. VIP Jr. VIP

    Joined:
    May 10, 2012
    Messages:
    5,035
    Likes Received:
    3,207
    Location:
    Europe
    Home Page:
    Now this is a pretty awesome case study! Especially the h2 tag is very interesting. Thank you for sharing!
     
  11. conrulez

    conrulez Power Member

    Joined:
    Dec 29, 2009
    Messages:
    539
    Likes Received:
    429
    Gender:
    Male
    Location:
    USA
    Very Nice!!! Thanks for sharing this!
     
  12. emerica1184

    emerica1184 Regular Member

    Joined:
    Jan 6, 2011
    Messages:
    349
    Likes Received:
    86
    This is awesome. Would love to see more stuff like this. You got my gears turning.
     
  13. ladyboyboom69

    ladyboyboom69 Regular Member

    Joined:
    Feb 15, 2016
    Messages:
    309
    Likes Received:
    104
    Occupation:
    Tony Stark
    Location:
    Your closet
    Thank you so much! This is very useful keep up the good work with Case studies like this
     
  14. bartosimpsonio

    bartosimpsonio Jr. VIP Jr. VIP Premium Member

    Joined:
    Mar 21, 2013
    Messages:
    12,774
    Likes Received:
    11,427
    Occupation:
    COINZ
    Location:
    BUYAH
    Home Page:
    Strongest correlation is Google +1's....their dead social network....sigh.
     
    • Thanks Thanks x 2
  15. validseo

    validseo Jr. VIP Jr. VIP Premium Member

    Joined:
    Jul 17, 2013
    Messages:
    917
    Likes Received:
    527
    Occupation:
    Professional SEO
    Location:
    Seattle, Wa
    Yeah. I guess desperate times call for desperate measures.
     
    • Thanks Thanks x 2
  16. bartosimpsonio

    bartosimpsonio Jr. VIP Jr. VIP Premium Member

    Joined:
    Mar 21, 2013
    Messages:
    12,774
    Likes Received:
    11,427
    Occupation:
    COINZ
    Location:
    BUYAH
    Home Page:
    And yet they don't want to make Plus any easier to use.

    I honestly can't understand how to use Google Plus. It's the most complicated thing in the history of complicated things.

    Twitter is so simple to use.... But Google's goal is not a social network, but a personal identification system. That's their whole problem right there.
     
    • Thanks Thanks x 1
  17. SellingIceCream

    SellingIceCream BANNED BANNED

    Joined:
    Feb 22, 2017
    Messages:
    86
    Likes Received:
    91
    Gender:
    Male
    @bartosimpsonio in a nutshell:

    1.Spread scary hat seo about Google dismantling your PBN's, and tracking your SEO clients/packages/strategies from their services like Gmail, Google docs/spreadsheets/adsense/analytics etc.

    2.When confronted with facts, start the one liner frenzy comebacks going full-scary-retard on the assailants, because nothing could be further from the truth than the bullshit @bartosimpsonio spreads.

    3.If attacks keep escalating, pull a @mickyfu esque reply , confronting the assailants with a dumb-enough question such as: "How many of you are there tonight, you must be {INSERT PREVIOUSLY BANNED USERNAME} with a brand new account"

    4.Rinse and repeat spreading more scary hat seo


    @bartosimpsonio intellect = protozoan
     
  18. kickthat

    kickthat Jr. VIP Jr. VIP

    Joined:
    Sep 18, 2014
    Messages:
    477
    Likes Received:
    598
    Gender:
    Male
    Location:
    UK
    • Thanks Thanks x 1
  19. kontentguy

    kontentguy Regular Member UnGagged Attendee

    Joined:
    Jun 10, 2014
    Messages:
    433
    Likes Received:
    142
    Gender:
    Male
    Occupation:
    White hat agency
    @validseo great stuff.

    followup thoughts:

    1) year in page title makes sense. i actually specifically avoid years in my urls so that they can be evergreen - wonder if thats connected at all, as something with 2016 in url probably started to drop in jan/feb of 2017 bc aint nobody searching for 2016 anymore

    2) https - great confirmation here

    3) h2s - i think bc many WP themes have the h1 automatically match the title tag (or the reverse), and h2s are deeper in the article, and indicate stronger density
     
  20. kickthat

    kickthat Jr. VIP Jr. VIP

    Joined:
    Sep 18, 2014
    Messages:
    477
    Likes Received:
    598
    Gender:
    Male
    Location:
    UK
    @validseo out of interest, have you done a correlation for sites with keywords meta tag in bing?