Ripped Off or Bad at Math???

1. 69

   lrohnerposted 13 years ago

   I was just checking up on one of my relatively new hubs and compared Analytics pageviews with AdSense impressions, and found that my AdSense share was shockingly short of 60%. It's relatively new hub with under 1,000 pageviews so far, so I didn't think much of it.
   
   I looked at stats for a more trafficked hub and still found that I fell short--only I'm not sure how much.
   
   According to AdSense (which don't equal the AdSense impressions reported by Analytics -- they're off by about 2,000 impressions), I only had 52% of impressions when comparing to Analytics' or HP's total pageviews number and 55% when comparing it to Analytics' unique pageviews number.
   
   I don't get it. And 55% is closer, but that's still over 5,000 pageviews that have disappeared into thin air.
   
   The closest thing I could find to the 60% is if I used the AdSense impressions reported by Analytics (which isn't what I got paid on) divided by the Unique Pageviews reported also by Analytics, it came out to 59% almost 890 pageviews short.
   
   My head hurts now. Anyone know what the heck is going on???

   replyreport
   1. 62

      Wayne Orvisburgposted 13 years agoin reply to this

      I thought HP took your clicks for 40% of the time. I'm probably mistaken though.

      replyreport
      1. 69

         lrohnerposted 13 years agoin reply to this

         Yup. HP takes 40% and we're supposed to get 60%. That's my point. I'm not getting 60% according to the stats.

         replyreport
2. 84

   psycheskinnerposted 13 years ago

   That stats vary widely for many reasons.  For example Blogger syas one blog of mine got about 10,000 hits last month.  Statcounter says about 6000, Project Wonderful and Adsense says about 7000.
   
   It comes down to when the time period is counted, what counts as a hit, what counts as a unique, when the crucial bit of the pages loads and all sorts of other stuff.  Often a page will load and the person will leave before the ad loads.  Stuff like that.

   replyreport
   1. 69

      lrohnerposted 13 years agoin reply to this

      Right. That's why I didn't think twice on a new hub with small numbers. I would expect to see discrepancies. But an older hub with significantly more traffic should have come close to evening out those oddities.
      
      I assume HP is counting impressions that mirror our hubs' traffic stats. I hadn't thought about people that might be using ad blocker services. If that is the cause of the discrepancy, that means that somewhere between 8% and 10% (5,000 to 7,000) of the people that visited that one hub had ads blocked or didn't wait for the ads to load. Does that number sound right? If it is, that's a scary prospect...

      replyreport
3. 87

   WryLiltposted 13 years ago

   I have no idea if this will help (the numbers you stated started to confuse me too!).
   
   Do you have Adsense & Analytics linked? If so go into your Content>Adsense in Analytics and at the bottom right of the stats mine says "0.59 AdSense Page Impressions / Visit".
   
   Not sure if that part would help (usually averages .60 to .59 so figured it had something to do with the share?)

   replyreport
   1. 69

      lrohnerposted 13 years agoin reply to this

      Genius, WryLilt! I don't usually track AdSense in Analytics, so I never noticed that before. My overall share is exactly 60%.
      
      But now I'm even MORE confused. Here's what I never realized before. That 60% of impressions that we get is not "impressions per hub," it's "impressions per account" -- or at least that's how it seems. Make sense?
      
      So I just discovered that my two highest AdSense-earning hubs  only got 50% and 54% of AdSense impressions respectively in October, while lower-converting hubs got more than 60% -- I saw one at 64%.
      
      Not a happy camper here. Not happy at all.

      replyreport
      1. 87

         WryLiltposted 13 years agoin reply to this

         Hmmm. Maybe watch them next month and see if those numbers swap around and see how random it really is.
         
         Glad I could help.

         replyreport
         1. 69

            lrohnerposted 13 years agoin reply to this

            I've gone back as far as January. My two consistently high-performing hubs have not gotten more than 55% share during that time -- and usually less than that. That's thousands of lost impressions for those hubs. *pouting*

            replyreport
4. 95

   wildernessposted 13 years ago

   Is it possible that slide shows are throwing the numbers off?  HP counts them separately, analytics seems to lump them into the count, and I'm not sure that adsense even counts them at all, although they report clicks from them.
   
   Looking at my totals on HP for all hubs all time, slide shows are about 20% of the total.

   replyreport
5. 71

   sunforgedposted 13 years ago

   Im not sure if i should admit this but Im 100% certain the impression share is per hub not per account.
   
   Ive manually checked the ad ids for ridiculous sample amounts many times. Its always works out appropriately.
   
   
   Adsense/analytics cant even get earnings or imnpressions down exactly - hence the entire graph dedicated to discrepancy in your analytics account.
   
   My account only show a .48 impression share - but thats because at least 15000 views a month do not have adsense enabled on the hubs.
   
   Do you have adsense turned off or disabled on any hubs?

   replyreport
   1. 69

      lrohnerposted 13 years agoin reply to this

      Nope.

      replyreport
6. 69

   lrohnerposted 13 years ago

   @SF - But the weird thing is, the ONLY number I could find that works out to 60% exactly is the number WryLilt referred to in Analytics. It's when I dug deeper that I saw that some of my Amazon-focused hubs are getting 62%, 64% or even 68% share, while my highest AdSense performers with consistent CTRs are getting 55% or less.
   
   Edited to add: Your method of sampling doesn't take into account those PVs that had ads disabled -- which I'm sure is at least some percentage of the views, and Analytics doesn't count those in their AdSense-impression numbers.

   replyreport