Page 2 of 2 First 12
  • Jump to page:
    #16
  1. Contributing User
    SEO Chat Adventurer (500 - 999 posts)

    Join Date
    Apr 2011
    Posts
    970
    Rep Power
    1550
    Again - the point of my thread was: with Invisible ReCaptcha is there general agreement to throw everything Google said before out the window and generally agree that user behaviour will be, if not already is, part of their algo now.

    As such what anyone said last year or in 2008 or whenever would not be relevant.

    I do disagree with the reliability of these signals. A high bounce rate is not 'bad' and a low bounce rate is not 'good'. These sort of things are judged comparing to your neighbours, not a static figure.

    Google has always made a great deal of user experience in Adwords. It means that you can bid lower and appear higher than a competitor. It has always seemed odd to me that their official line is "We do it there but we don't do it here".

    This aside my experience has been that improving user behaviour signals improves rankings. Now it could be that while I'm aiming to improve these signals I'm actually doing something else and that is what is causing the improved rankings. From a client perspective they don't really mind.
  2. #17
  3. No Profile Picture
    Contributing User
    SEO Chat Discoverer (100 - 499 posts)

    Join Date
    Mar 2017
    Location
    Sydney, Australia
    Posts
    317
    Rep Power
    639
    Hi Doodled,
    I would like to put an exercise in logic to the SEOs about when/if engagement signals are/may become part of G's ranking signals.

    I've arranged it around three SE issues:

    1. Why SEs rank individual words?
    2. The SE indexing process.
    3. Google Adwords vs generic search

    1. Why SEs rank score individual words?
    When G was first launched (circa 2000) it was geared to indexing 100 million pages. Back then, good search results could be attained with an average two word search. Now G has indexed 130 trillion pages and the average search has grown to four words.

    As the length of search queries lengthen, four things usually ensue. More search words =

    • There are fewer pages that match the search result
    • Fewer ranking points are needed to rank top of SERPs that have less matching pages
    • There are more words across which ranking points can be scored.
    • The more words that can be targeted for ranking points, the more variable are the potential SEO tactics

    The fact that searchers have needed to progressively lengthen their search queries to obtain relevant results, to me is a big hint that "engagement" signals can't create enough relevancy signals to achieve "good" results.


    2. The SE indexing process
    There is a G algo sort sequence involved in preparing SERP lists. This is based on G's original description in 1997 and its current description of "How Search Works".

    This is how I understand G processes its search sequence after it has indexed a web page file.

    Step 1: Possible modification of original search query words by RankBrain
    Step 2: Create a list of pages that qualify for the RB modified search query
    Step 3: Analysis of the individual words in the search query
    Step 4: Selective implementation of special algo factors. (Eg Location, query deserves freshness (QDF), mobile search, https, mobile formatted pages, etc.)
    Step 5: If there are any "engagement" signal factors (I don't believe there are any) IMHO, their importance would need to be relegated to the very end of the ranking process.

    A problem with "engagement" signals (CTR, bounce rates, dwell time, time on page, etc.) is that they do not contribute to a SE's understanding of the words in a search query, unless only search query related "engagement" signals are used and, I suggest this would likely bring huge costs for little return.

    3. Google Adwords vs generic search
    G has repeatedly said that there is no connection between Adwords and its generic ranking results.

    If an SEO works in the competitive USA market, he/she may not see how clear the distinction is vs. a picture from smaller country search markets.

    Main points of difference between Google Adwords and generic search results starts with:

    Step 1: Create a list of ads
    G ads are NOT web pages. Qualifying for the Adwords results list is based on the selection of one or more words (+any other words).

    Step 2: The cost per click that was bid by the advertiser.


    Step 3: The size of the advertiser's budget


    Step 4: The quality of the advertiser's landing page.
    Sorry, but I don't see any factors of relevance between generic search results and paid ad rankings.

    Doodled, You ask specifically about Invisible ReCaptcha. My first question is, how widely used is it? If you read G's pronouncements, it says that if a metric is not widely used on the web, it will not be used in its ranking signal.

    So, how about we discuss SE ranking logic and based on that, discuss where engagement signals may play a part?
  4. #18
  5. SEO Since 97
    SEO Chat Mastermind (5000+ posts)

    Join Date
    Mar 2011
    Location
    Arizona
    Posts
    7,735
    Rep Power
    4753
    The main ingredients are good content, good site structure and good IBL. The rest is minabable.
    Last edited by Test-ok; Aug 7th, 2017 at 05:11 AM.
  6. #19
  7. No Profile Picture
    Contributing User
    SEO Chat Discoverer (100 - 499 posts)

    Join Date
    Mar 2017
    Location
    Sydney, Australia
    Posts
    317
    Rep Power
    639
    Hi Test-OK,
    No doubt, good content and site structure are important SEO issues.

    However, it all starts with G deciding what pages qualify for the SERP list. With 4 words being the average search length now, page content for 4+ word search queries needs to be the starting point for SEO, IMHO.

    I was hoping folk would delve deeper into what factors G uses to decide which pages as diverse as a long list of products, recipes, services or business categories vs. pages of long, good content about a single product, recipe, service or business should rank top.

    I understand, the issue raised by the OP in this post is whether "engagement" signals can play a significant part in the ranking process.

    Do you think "engagement" signals can play a major part in ranking SERPs?
  8. #20
  9. No Profile Picture
    Registered User
    SEO Chat Explorer (0 - 99 posts)

    Join Date
    Mar 2014
    Posts
    24
    Rep Power
    0
    Guys, Google don't use your analytics reports in ranking algo, that's true, because if they do, I personally don't use it for privacy reason.

    Also site like wikipedia/reddit not using G analytics, so they don't depend on it at all.

    But Google can create their own metrics same like analytics to check weather User is satisfied or not.
    Last edited by Arjun3315; Aug 8th, 2017 at 04:50 AM.
  10. #21
  11. SEO Since 97
    SEO Chat Mastermind (5000+ posts)

    Join Date
    Mar 2011
    Location
    Arizona
    Posts
    7,735
    Rep Power
    4753
    Engagement, it can and maybe not I believe traffic from many ip's with time spent is a plus factor but it can't be proven extensivly either way.
    Last edited by Test-ok; Aug 8th, 2017 at 04:55 AM.
  12. #22
  13. Digital Marketing
    SEO Chat Mastermind (5000+ posts)

    Join Date
    Aug 2008
    Location
    Florida
    Posts
    7,342
    Rep Power
    5031
    Originally Posted by JohnAimit
    Hi Test-OK,
    No doubt, good content and site structure are important SEO issues.

    However, it all starts with G deciding what pages qualify for the SERP list. With 4 words being the average search length now, page content for 4+ word search queries needs to be the starting point for SEO, IMHO.

    I was hoping folk would delve deeper into what factors G uses to decide which pages as diverse as a long list of products, recipes, services or business categories vs. pages of long, good content about a single product, recipe, service or business should rank top.

    I understand, the issue raised by the OP in this post is whether "engagement" signals can play a significant part in the ranking process.

    Do you think "engagement" signals can play a major part in ranking SERPs?
    Hey John, I like the path you and Tom are on. Would you like to start a new thread maybe listing the elements of SEO highest to lowest?
  14. #23
  15. No Profile Picture
    Contributing User
    SEO Chat Discoverer (100 - 499 posts)

    Join Date
    Mar 2017
    Location
    Sydney, Australia
    Posts
    317
    Rep Power
    639
    Hi KP,
    If I did, the post would probably have a name like, The Fallacies of Top 200 SEO Ranking Signal Lists.

    Could you make positive contributions to something like that?
  16. #24
  17. Digital Marketing
    SEO Chat Mastermind (5000+ posts)

    Join Date
    Aug 2008
    Location
    Florida
    Posts
    7,342
    Rep Power
    5031
    that already sounds pretentious, probably not. How about top 5 elements of local seo?
  18. #25
  19. Contributing User
    SEO Chat Adventurer (500 - 999 posts)

    Join Date
    Apr 2011
    Posts
    970
    Rep Power
    1550
    Originally Posted by JohnAimit
    The fact that searchers have needed to progressively lengthen their search queries to obtain relevant results, to me is a big hint that "engagement" signals can't create enough relevancy signals to achieve "good" results.
    Possibly so, as said by others, difficult to conclude/prove.

    Originally Posted by JohnAimit
    If there are any "engagement" signal factors (I don't believe there are any) IMHO, their importance would need to be relegated to the very end of the ranking process
    Again, possibly so. I am not arguing that engagement signals could become the signal, but new technology that Google has achieved means they may now become a signal (although I believe they already are but that is an aside)

    Originally Posted by JohnAimit
    Sorry, but I don't see any factors of relevance between generic search results and paid ad rankings.
    I think you misunderstand me but first you missed out of your list 'Ad relevancy) or basically CTR. I was not saying they were connected (although I'll come to that and say it in a minute!). I was saying Google uses these signals in Adwords but then publicly says it can't use them in organic ... which I find odd.

    Now I will say that they are connected! Again I see this a great deal in my experience although I can't put my finger on which exact two factors correlate but I also approach this from an SE engineers point of view.

    Let's say the search term "Cheese Cracker ideas" and say the top 5 ranking pages have CTR of 50%, 40%, 30%, 20%, 10% and they all have a bounce rate of 60%.

    Someone starts an Adwords campaign that targets this keyword. Every time their ad is shown 70% of users click on it and never bounce (for arguments sake). As an SE engineer should I totally ignore that or should I be saying, "here is a signal that this landing page behind the ad is more relevant to the searches or 'cheese crackers ideas' than anything else in the SERPs, it would improve the SERPs is it were there organically ....."

    The trouble is to believe in that second part you must believe in the first part. If you don't they cannot be connected by definition - because Google is not using engagement signals it cannot compare engagement signals of organic with adwords ...

    Originally Posted by JohnAimit
    You ask specifically about Invisible ReCaptcha. My first question is, how widely used is it?
    This was not so much about the usage of Invisible ReCaptcha but the technology behind it being applied to Google's own search results so they know when humans are clicking in the SERPs and when a bot is running around trying to game things.
  20. #26
  21. No Profile Picture
    Contributing User
    SEO Chat Explorer (0 - 99 posts)

    Join Date
    May 2017
    Posts
    23
    Rep Power
    0
    If you're search behavior is unusual, Google will still bring up a page asking if you're a robot. They used to give users a regular captcha or, if JavaScript was on, a ReCaptcha checkbox. They eliminated the regular captcha. That's all. :/

    Got stuck on one of these "prove you're human" pages last week. Curse me for trying to use advanced search features with client-side scripting turned off (cellphone, limited bandwidth,) and for looking at more than the 1st page. Hey, it's not my fault nothing looked interesting for 5 pages. If you want a screen-cap I took one; I'll just have to find it/figure out how to upload it.
Page 2 of 2 First 12
  • Jump to page:

Similar Threads

  1. Typical Bounce Rate for those with High Rankings?
    By ericwoo in forum Google Optimization
    Replies: 14
    Last Post: Dec 28th, 2012, 12:54 AM
  2. Google's Buy of reCAPTCHA Hurt Internet Security? (PC World)
    By RSS_News_User in forum Technology News
    Replies: 0
    Last Post: Sep 17th, 2009, 01:03 PM
  3. Google Will Buy reCAPTCHA To Help Scan Books (NewsFactor)
    By RSS_News_User in forum Technology News
    Replies: 0
    Last Post: Sep 17th, 2009, 12:04 PM
  4. Making Page Invisible on Google (due to error)
    By mbot in forum Suggestions & Feedback
    Replies: 0
    Last Post: May 13th, 2005, 11:50 AM
  5. Invisible in Google - Why?
    By timothy911 in forum Google Optimization
    Replies: 12
    Last Post: Jun 7th, 2004, 03:15 AM

IMN logo majestic logo threadwatch logo seochat tools logo