16 {{ upvoteCount | shortNum }}
16 {{ upvoteCount | shortNum }}

Craziest Thing I have ever seen in WMT

by maddogg271
14 replies
  • SEO
    • |
Hey Guys,

I am new to the forums, but I am looking for feedback to what reasons could be causing extreme fluctuations in traffic in correlation with indexed pages in WMT. This is a very large website with traffic in excess of 11k uniques per day, literally overnight traffic dropped by 85%. The only correlation that occurred was that we simply disallowed ia_archiver in robots.txt.

Before looking at the pics below:
- We confirmed no technical errors occurred
- Sitemaps are solid
- We have received no manual penalties or actions
- No significant errors or access denied issues


30 Day Sitemap





Average Site Position according to WMT





Clicks in WMT





Crawl Stats





Google Analytics


#craziest #thing #wmt
  • Profile picture of the author savidge4
    Originally Posted by maddogg271 View Post

    Hey Guys,

    The only correlation that occurred was that we simply disallowed ia_archiver in robots.txt.
    The only indicator of any kind I can see would be the page load time. I might think you have a page load issue of some kind? ( partial page load vs full page )

    I would suggest getting in and doing some region specific proxy server based searches and see if there is an issue somewhere.

    I might even go so far with injecting a "Cant See this page?" popup at the very top of the page for a day or so and see how many clicks its getting?
    Signature
    Success is an ACT not an idea
    {{ DiscussionBoard.errors[10180012].message }}
    • Profile picture of the author maddogg271
      The only indicator of any kind I can see would be the page load time. I might think you have a page load issue of some kind? ( partial page load vs full page )

      I would suggest getting in and doing some region specific proxy server based searches and see if there is an issue somewhere.

      I might even go so far with injecting a "Cant See this page?" popup at the very top of the page for a day or so and see how many clicks its getting?
      Thanks for the suggestion!! Our page speed insights have dramatically improved over the past couple months as we had to continue to optimize for traffic.

      It is something I am completely stumped at.
      {{ DiscussionBoard.errors[10180153].message }}
      • Profile picture of the author savidge4
        Originally Posted by maddogg271 View Post

        Thanks for the suggestion!! Our page speed insights have dramatically improved over the past couple months as we had to continue to optimize for traffic.

        It is something I am completely stumped at.
        I would say that your page speed has increase to the point of not possible. It appears you may be below a 600 or so, looks more 300 ish... Yeah if the server is in the other room maybe.. but site wide and visitor wide... seriously not a chance.

        The other question is your traffic breakdown mobile to desktop? did you lose one or the other? Or did they both drop equally?
        Signature
        Success is an ACT not an idea
        {{ DiscussionBoard.errors[10180251].message }}
        • Profile picture of the author maddogg271
          Originally Posted by savidge4 View Post

          I would say that your page speed has increase to the point of not possible. It appears you may be below a 600 or so, looks more 300 ish... Yeah if the server is in the other room maybe.. but site wide and visitor wide... seriously not a chance.

          The other question is your traffic breakdown mobile to desktop? did you lose one or the other? Or did they both drop equally?
          They both equally dropped, here's a graph showing the mobile vs desktop drop. You have a really good point, the speed is on average 150ms - 225ms, that seems to be a near impossibility. I wonder if google started deranking most of our pages due to lack of rendered content on the page?

          {{ DiscussionBoard.errors[10180351].message }}
          • Profile picture of the author maddogg271
            I also wanted to mention that in WMT, the page load time is for the actual non-rendered source, so 150ms - 300ms isn't actually all the unreasonable I do not think.
            {{ DiscussionBoard.errors[10180389].message }}
            • Profile picture of the author savidge4
              Originally Posted by maddogg271 View Post

              They both equally dropped, here's a graph showing the mobile vs desktop drop. You have a really good point, the speed is on average 150ms - 225ms, that seems to be a near impossibility. I wonder if google started deranking most of our pages due to lack of rendered content on the page?
              No content no base for rank... it would be like suggesting a blank page would rank for a particular keyword terms. ( this blank page is for keyword A and this other blank page is for keyword B )

              Originally Posted by maddogg271 View Post

              I also wanted to mention that in WMT, the page load time is for the actual non-rendered source, so 150ms - 300ms isn't actually all the unreasonable I do not think.
              I would suggest reading this: Google Analytics Site Speed vs. Google Webmaster Tools Site Performance Explained – Guy's Pod to get a better idea of what is measured and how. They are measuring Onload to complete, so your current load times I would say are way out of whack and of great concern.

              The other thing that came to mind is did you switch to a dynamic page load by any chance? I guess the terms is "Lazy Loading" but check this out to see if it is of value: https://productforums.google.com/for...rs/aMLMOLzLWhQ

              Other than that I would really start breaking down traffic by say country and see if there is a correlation between distance and load times. Is it longer loading in the states and much faster abroad? ( I am assuming you have US based host servers )
              Signature
              Success is an ACT not an idea
              {{ DiscussionBoard.errors[10180413].message }}
  • Profile picture of the author mkgg
    SiteSpeed won't cause a dramatic decline like this. Most likely cause is either a negative seo attack or the link-building that you did yourself.

    Or other sites have emerged that are taking your position.
    {{ DiscussionBoard.errors[10180419].message }}
    • Profile picture of the author savidge4
      Originally Posted by mkgg View Post

      SiteSpeed won't cause a dramatic decline like this. Most likely cause is either a negative seo attack or the link-building that you did yourself.

      Or other sites have emerged that are taking your position.

      There were over ONE AND A HALF MILLION pages indexed... I don't think someone just strolled in and bounced 500,000 pages. ROFLMAO
      Signature
      Success is an ACT not an idea
      {{ DiscussionBoard.errors[10180427].message }}
      • Profile picture of the author mkgg
        Originally Posted by savidge4 View Post

        There were over ONE AND A HALF MILLION pages indexed... I don't think someone just strolled in and bounced 500,000 pages. ROFLMAO
        Big Deal, any natural non-affiliate junk site have thousands of pages and they don't even rank. The number of indexed pages has nothing to do with rank. My forums tanked because of a negative seo and i had close to a million pages indexed, still do but the ranking tanked.
        {{ DiscussionBoard.errors[10180445].message }}
        • Profile picture of the author savidge4
          Originally Posted by mkgg View Post

          Big Deal, any natural non-affiliate junk site have thousands of pages and they don't even rank. The number of indexed pages has nothing to do with rank. My forums tanked because of a negative seo and i had close to a million pages indexed, still do but the ranking tanked.

          But that is the point here... we are not talking about loss of rank.. we are talking DE INDEXED, and in large numbers... Negative SEO would be like a silent creeper in the night over time things would drop.. a page here a page there then more progressive to the point of crippling... THIS appears to have been CLICK and done.
          Signature
          Success is an ACT not an idea
          {{ DiscussionBoard.errors[10182014].message }}
      • Profile picture of the author nettiapina
        Savidge offered some good suggestions.

        You mentioned what you see as possible correlations, but have there been changes to the site code or hosting platform? Do you know of any technical challenges that the host or server hotel might be facing?

        The loading times indeed seem abnormal, and there's also a drop in the number of indexed pages. The root cause could be same.
        Signature
        Links in signature will not help your SEO. Not on this site, and not on any other forum.
        Who told me this? An ex Google web spam engineer.

        What's your excuse?
        {{ DiscussionBoard.errors[10180451].message }}
        • Profile picture of the author maddogg271
          With more investigating, I think we may have found the problem. We are not exactly sure. We decided to hide one table, with content, that exists on nearly every page on the website from GoogleBot, however we when a human user hit the page they would see it.

          Even though it wasn't malicious in anyway, I am wondering if Google considered this Cloaking and started to deindex pages rapidly.

          What do you guys think?
          {{ DiscussionBoard.errors[10181305].message }}
          • Profile picture of the author savidge4
            Originally Posted by maddogg271 View Post

            With more investigating, I think we may have found the problem. We are not exactly sure. We decided to hide one table, with content, that exists on nearly every page on the website from GoogleBot, however we when a human user hit the page they would see it.

            Even though it wasn't malicious in anyway, I am wondering if Google considered this Cloaking and started to deindex pages rapidly.

            What do you guys think?
            There would be factors to determine cloaking or not... Just because you say don't look.. they do, but then here is the determination on their side as to the nature of the hide. Malicious or not.

            IF the content is visible to the user I would wonder what the thought was to hide in the first place? - if it is the same content repeated over and over introducing an image as a replacement I think would make greater sense? ( Less server load over time and a reduction in page code ) BUT, if each page is hitting on a set of separate keywords and the table is in context to the information I would without question leave it.

            The other side of this is how large is the table? what percentage of content are you knocking out with hiding this?

            For giggles at this point, I might try re introducing it to see what happens.

            But, this still in no way explains the extreme drop in load times. - I would be real concerned with that
            Signature
            Success is an ACT not an idea
            {{ DiscussionBoard.errors[10182007].message }}
            • Profile picture of the author maddogg271
              The table was variable information, however we didn't want Google to crawl the info for Meta information, we wanted it to be visible to the user, but didn't want google to index the content on the table.
              {{ DiscussionBoard.errors[10184740].message }}

Trending Topics