Who is right - Analytics or HubPages?

1. 85

   Pcunixposted 14 years ago

   I haven't paid a lot of attention to Analytics on HubPages because I'm just too new here - not enough page views and not enough elapsed time.  I do use Analytics extensively at my much older and much higher traffic main site.
   
   My attention was attracted to a burst of activity in the last week at one particular hub.  I say "burst", but really it's just a micro burst - a few hundred views out of the blue, which is something I ordinarily wouldn't even bother looking at - too small to mean anything.
   
   But I did look into this and was surprised to find a big discrepancy between what Analytics tells me and what HP "stats" say.  Analytics says that most of those views (78%) were direct entrances and that most of that was by way of Google.  HP stats says the opposite - 80% from Hubpages and the rest indiscernable.
   
   It seems odd to me that there should be such a variance.  At my main site, I have distrusted Google at times and have gone to raw logs to resolve the discrepancy, but of course that's not an option here (too bad that it is not, by the way - it would be simple enough for HP to provide raw logs on request).
   
   Has anyone else ever noticed conflicting stats?

   replyreport
   1. 0

      Twenty One Daysposted 14 years agoin reply to this

      Sure there is a variance.
      HP is a "random" page view of around 60% of 100%. Which translates to 60% [ which is actually 60%  x .75, as Google takes a cut of the impression]. What I have noticed at times, a page view does not qualify on Google because it was an HP view aka Indiscernible. Also noted is when you view/edit mode a page it shows on HP as a page view until the system flushes (once an hour or once a day @ 6am PST).
      
      An "Indiscernible Source" just means the javascript HP is using does not or can not define/store the data, as the script is not written to do so. Ergo, the variable.
      
      Oddly, HP will crow high numbers on my "daily" views (which is a continuous clock of 24 hours v. a stop-at-midnight 24 hour cycle (like Google).
      
      The "random function" for "60/40" share is where the issue appears to begin and end. In between is the clock and data variables...
      
      EDIT:
      Also something to remember: in your stats under views, if it reads HubPages as the source of the view, 99% of the time, it is a "logged in member". This will show as a page view but not necessarily a Google page impression. I have 75 articles-in-channel on Google which show a massive gap between HP numbers and Google numbers.

      replyreport
      1. 85

         Pcunixposted 14 years agoin reply to this

         I forgot about the 60/40, but I am not sure that can explain what I see here.  I  did take the rolling clock into account.
         
         I need to think about the 60/40 - it could be responsible. I need to work on a larger data set over longer time.

         replyreport
         1. 0

            Twenty One Daysposted 14 years agoin reply to this

            60/40 LESS logged in views by HubPages members...
            So if you have 50 HP views and 10 Google search views, 10 x .60 = 6 would be your actual impression, even though HP is showing 60 views in total.
            
            Even though the Ads show up, when your logged in, the analytics is not necessarily passing the data as a 'clean' view to google even though HP is recording a HubView.
            
            I found Kontera somewhat helpful with tracking impressions, but the trouble is they do not show the source/page url, unless there is a click. In which case they record the url and keywords. Plus, it is also random, as Vibrant intelliTXT appears in place of Kontera.
            
            EDIT:
            Give you an example. Today my article on Red Bull (which is my highest viewed article ever) is showing 10 views. 5 of those views are from Google engines; 5 from HP. In my AdSense account it shows 3 views ( 5 x .60 = 3 ).

            replyreport
            1. 85

               Pcunixposted 14 years agoin reply to this

               Well, again, that doesn't explain what is happening here.  This isn't a different number of views, this is an inversion of entrance sources.
               
               Also, if you look at page source for a hubbers page, they have something like:
               
               gaq.push(
               ['t1._setAccount', 'UA-69824-6'],
               ['t1._trackPageview'],
               ['t2._setAccount', 'UA-286186-1'],
               ['t2._trackPageview']
               
               which means that both their Analytics account and the hubber gets the page view tracked, which makes perfect sense.

               replyreport
2. 0

   BRIAN SLATERposted 14 years ago

   like you pcunix i have a analytics account but have been informed that not enough lapsed time has occured yet- but i have noticed on some of my hubs it says the traffic has come from an undescernable(??) source- but when i have checked analytics it has told me from where the traffic has come from. Why can't Hp's.

   replyreport
   1. 85

      Pcunixposted 14 years agoin reply to this

      Keep in mind that Analytics gets its info from Javascript on the page. HP is probably analyzing web logs.
      
      Discrepancies I understand, but polar opposites I do not.

      replyreport
3. 61

   idreesfarooqposted 14 years ago

   I have both account at analytic and hubpages. I always find a great difference between Analytics page views and hubpages page views. I am still trying to figure out what is the actual reason. Still not get one, perhaps one of top hubber here will answer this.

   replyreport
4. 80

   Benjimesterposted 14 years ago

   Yeah, I've noticed a growing discrepancy too.  It's pretty small, but always there.  Google usually reports slightly less overall traffic than HP does, which is a bit disturbing because, like you all said, those phantom view that HP says are from an "Undiscernable Source" are kind of strange.  It gets me to wondering whether they really exist or not?
   
   On the other side of the coin, it's important to note that Google Analytics and HP Analytics report views at different intervals.  Google displays its findings at the end of the 24 hour period, whereas HP is continually updating round the clock.  So you should probably be seeing some level of discrepancy because of that I would think.

   replyreport
5. 75

   WryLiltposted 14 years ago

   I have no idea if this will help or not but here goes:
   
   Did you:
   
   Put a new link to another hub a well trafficked hub?
   Put a new RSS feed in?
   Have that hub for some reason rank higher in the RSS feed?
   
   
   I'm just asking because I've seen views that follow a link (for instance they visit hub A from google, and are redirected from there to hub B.)
   When the person reaches hub B, that view is counted as a hubpages view, even though originally they came from google. I'm not sure if analytics would do the same.

   replyreport
   1. 85

      Pcunixposted 14 years agoin reply to this

      That could be.  It may be that Google and HP track sources differently.

      replyreport