Forum Moderators: martinibuster

Message Too Old, No Replies

Adsense Stats Different From Webtrends and Analytics

Webtrends displays much higher results

         

sc123

1:24 pm on May 31, 2006 (gmt 0)

10+ Year Member



Here are a few numbers to show the difference between what Adsense is reporting and what Webtrends and Analytics are reporting. The Websense number is from the line that reads "Page Views (Impressions)". The Analytics number is "Visits".

For yesterday, May 30th:
Webtrends: 385
Adsense: 24
Analytics: 12

For the month of May:
Webtrends: 4,867
Adsense: 449
Analytics: 382

I can't seem to explain the massive difference. I was hoping someone here could shed some light on the differences. Adsense code is one pretty much every page that a visitor would see, although Analytics is only on the home page (since pretty this is the only page people visit anyway).
Thanks,
SC

carguy84

8:19 pm on May 31, 2006 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



One(Adsense) is tracking the number of page views your site gets.

The others are tracking the number of users who go to your site (visits).

So it looks like, on average, a visitor views 10 pages per visit, which is good.

Chip-

SEO_Report

8:45 pm on May 31, 2006 (gmt 0)

10+ Year Member




One(Adsense) is tracking the number of page views your site gets.
The others are tracking the number of users who go to your site (visits).

So it looks like, on average, a visitor views 10 pages per visit, which is good.

So, 385 visitors (Webtrends) generated 24 pv (Adsense)!

Numbers simply don't match!

I am seeing the same trend. According to my logs, PSA account for less than 1%, and 98% of my visitors have javascript enabled, yet G shows 10 times less pageviews than the real figures!

hunderdown

8:53 pm on May 31, 2006 (gmt 0)



There have been some other discussions about this recently. The gist of them is that you shouldn't assume that AdSense is undercounting. AdSense doesn't count search engine bot visits, which many stat programs do. If WebTrends is one of the "overcounters," that could explain the discrepancy, particularly for a site that isn't getting a lot of visitors.

loganz

9:00 pm on May 31, 2006 (gmt 0)

10+ Year Member



youre also forgetting that some people actually block adsense ads via adblockers. its probably not a significant amount of people but numbers will always be different between different counters/trackers

kaz

9:35 pm on May 31, 2006 (gmt 0)

10+ Year Member



does anyone have a stats program that is consistently in sync with adsense?

Content_ed

10:08 pm on May 31, 2006 (gmt 0)

10+ Year Member



You should be able to get a Page Views number out of Analytics as well for comparison with Adsense. We've found that Analytics does a much better job with stats than our server package, which counts endless bot visits the same as real visitors.

21_blue

10:49 pm on May 31, 2006 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



sc123 wreote:
For yesterday, May 30th:
Webtrends: 385
Adsense: 24
Analytics: 12

For the month of May:
Webtrends: 4,867
Adsense: 449
Analytics: 382

I can't seem to explain the massive difference.

This type of topic has come up many times before, and there is usually a simple explanation that confirms Adsense stats are reliable. So, working on that assumption...

The difference between Adsense and Analytics (I presume Google analytics) suggests to me either:

  • that the Analytics code may be missing off some pages or
  • that you are counting adsense ad impressions rather than pages.
Analytics should report the same or slightly higher than Adsense (it does for our sites - the difference being due, for example, to the lack of any relevant ads to show).

I don't know what Webtrends stats you are looking at, but the massive differences between Webtrends and the other stats suggest either:

  • that you are looking at Webtrends hits rather than page views or
  • that you are being visited more by bots than real people (Webtrends may be counting all the bots, which only retrieve the basic page content, but Adsense and Analytics only count real people who execute the javascript)
I use Mach5 analyzer which has filters to remove all bots and my own visits, and all the stats correlate with each other more or less (and with a count of a gif file that appears on all pages).

sc123

1:34 am on Jun 1, 2006 (gmt 0)

10+ Year Member



carguy84: As I mentioned, I posted Webtrends "Page Views", not "Visits".

21_blue: I can understand why Webtrends might be higher due to it counting bots and every single file hit (which Analytics and AdSense do not), but I don't understand why there would be such a difference between AdSense and Analytics, especially since the code for each are on the same pages and only those pages. Again, the Webtrends data is Page Views, not hits or visits. Just to clarify, here is what I'm getting from Webtrends in "Hits":

Webtrends Hits on May 31st: 2,873
Webtrends Hits in May: 50,879

I'd really like to figure this out as my ads are performing miserably (no one has clicked on yet)!

Thanks,
SC

21_blue

9:52 am on Jun 1, 2006 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



21_blue: I can understand why Webtrends might be higher due to it counting bots

You can verify this with another stats package. Download your log file and a demo copy of Mach5 Analyzer, and select an automatic filter to remove all bots/spiders.

I don't understand why there would be such a difference between AdSense and Analytics, especially since the code for each are on the same pages and only those pages.

It is worth double-checking there are no "typos" in the code on your local disk, by using a multifile find utility to count how many times each piece of code appears.

Another source of discrepancy might be any filters that you have set in analytics.

Also, where is the code on the page - eg: if you have big pages with adsense at the top and analytics at the bottom, and the page doesn't finish loading, that might result in discrepancies. A lot of pages fail to finish loading. You can check how long the page will take to load on slow links using a utility such as is contained in the document properties section of the Hotdog html editor. Alternatively, try putting the analytics code just above the adsense code and see if it makes a difference to the discrepancies.

I also wonder if cache issues are causing the discrepancy somehow.

I'd really like to figure this out as my ads are performing miserably (no one has clicked on yet)!

I'm not sure resolving these discrepancies are going to help you improve ad performance. The adsense stats will be accurate. You probably need to look at other issues such as ad position (see the Google heat map) and page design to improve CTR or section targetting to improve ad relevance.

Rusty_P

9:38 pm on Jun 20, 2006 (gmt 0)

10+ Year Member



I complained to Google six months ago (and this week again).

Their answer was useless, that people disable "images, java and cookies" so you can account for those pages.

But that's hard to believe, my readers are not so advanced.

And it's getting worse everyday (BTW, forget about timezones and multiple ads).

For example, last sunday:

Adsense: 4000 impressions
Analytics: 2300 pages

Yesterday:

Adsense: 6180 impressions
Analytics: 4700 pages

And the other weekend was OK, so to me it's plain simple that something's rotten.

It's up to Google to admit it.

Rusty_P

9:50 pm on Jun 20, 2006 (gmt 0)

10+ Year Member



SORRY

Their answer was useless, that people disable "images, java and cookies" so you CAN'T add those pages.