Forum Moderators: martinibuster
For me this week has been a typical post 5th May week however yesterday I sent out my monthly newsletter to 18,000 subscribers and I have monitored it very, very carefully.
Fact, bandwidth up 32%, page views up 52%.
According to Adsense I had a completely standard day! In fact, their statistics show less page views than on Monday but 4% more ad impressions.
These extra ad impressions do not represent a 2% increase in pages, and before anyone says it there are more than enough ads for my industry for fresh impressions.
Unless I get a massive click dump today, now tell me there's nothing wrong with MY Adsense.
Who knows? Yours may be the same.
I did the same as you - sent out a newsletter - traffic to my site increased 45% the next day, 36% the following day. AdSense showed a decrease in traffic both days.
Really sad how these guys sold out.
Do google log PSA's? Could you be displaying a lot of those?
TJ
In any case, not related to this particular issue. For what it's worth, though I don't track pages as carefully, the general pattern of ebbs and flows for my site's traffic matches up with the AdSense reporting of impressions....
Insofar as PSA's are concerned, apart from when I first included Adsense about 18 months ago, I have never seen one on any page.
One thing is certain for today though, so far my eCPM is higher than I've ever seen before and if, note IF, this were to continue then my estimates would be that all which was lost yesterday would be recouped today.
Several others have commented on a disastrous Thursday which seems to be continuing into Friday:
[webmasterworld.com...]
[webmasterworld.com...]
there is a massive difference betweeen Google's statistics and mine.
Clearly there is something wrong there. PSA's was the only explanation I could think of.
Have you emailed AS support? Did you get a stock answer?
TJ
Also, AdSense ads are geotargeted, which means that PSAs could be showing up in Timbuktu even if they aren't displaying in Toledo.
What does AdSense Support have to say about the discrepancy?
If you want to find out if you're REALLY serving PSA's, whether you're seeing them or not, it's trvial. I don't know if PSA's are counted in the stats but I don't think so.
Install the collapsing ad unit code from Alternate Ads:
[google.com...]
Then you'll see something like this in your STATS showing how many times it loaded:
136791 /ads/google_adsense_script.html
Yes, I served 136,791 collapsing ads blocks instead of PSAs so I know where my stats were off. Also consider that bots crawling your site don't run adsense, nor do paranoid people with javascript disabled, and very possibly Norton Firewall whacks ad displays as well.
You can point all the fingers at AdSense you want but I'm getting more impressions from them than I am from affiliate programs, the ad blockers are getting ugly.
Another way to test it would be to install a small javascript test alongside adsense that loads from the server and then compare how many times your javascript executed vs. adsense PAGE impressions + alternate ad impressions and see what you come up with.
OK villagers, back to your witch hunt.
Install the collapsing ad unit code from Alternate Ads:
Useful tip - thanks Bill.
Have you emailed AS support? Did you get a stock answer?
Stock answer, no matter how many times I give them my statistics all I get is a canned response. Appalling customer service.
are you excluding crawlers and other bots from your server logs?
Yep, all filtered out.
Another way to test it would be to install a small javascript test alongside adsense that loads from the server and then compare how many times your javascript executed vs. adsense PAGE impressions + alternate ad impressions and see what you come up with.
Already have that and this is from where all the discrepancies have appeared and especially so yesterday.
I was prepared to give Google the benefit of the doubt and even prepared to acknowledge that maybe the advertising etc in my industry is slightly seasonal. Since the international trade is very, very busy and there is no point in advertising when one is already full of orders, however until checking all the figures this morning and seeing the huge disparity between Google's statistics and my own, well, that's just confirmed it for me that nothing with my Adsense earnings has been right since 05/05/05.
It's pages, it's clicks, it's the EPC, it's the lot and I am not the only one who can identify 5th May as the commencing date for all these discrepancies.
It's pages, it's clicks, it's the EPC, it's the lot and I am not the only one who can identify 5th May as the commencing date for all these discrepancies.
So you actually have a javascript loading something running alongside adsense just to compare against?
I would recommend you contact Google then, might take a little persistance to move up the food chain to someone that can help, but someone can probably help you sort it all out. I had some similar issues and they were extremely helpful, pointed out some stuff I didn't even realize was happening.
Right!
Turnover point - absolutely and definetely.
I was trying to patch my site after this date
but still see drop.
Many forms of advertising have been tried, and will be tried in this world of sin and woe. No one pretends that AdSense is perfect or all-wise. Indeed, it has been said that AdSense is the worst form of advertising except all those other forms that have been tried from time to time.
I am not the only one who can identify 5th May as the commencing date for all these discrepancies.
Well, about that date, I believe in fact it's only a "Show data by Page / Individual Ad Unit" thing. Since 5th May we can see a much greater difference between these two statistics, at least if we use an ad links unit in addition to normal ads on the same pages.
Surely they implemented something that date. See it by yourself, open two browser windows or tabs and look around May 5 with the two options "page" and "individual ad unit".
That's all about May 5, don't worry. I don't think it affects earnings at all.
Others have pointed out several common causes of differences among statistics (for example, if people set high security on IE browser -JavaScript disabled-, then they don't see AdSense ads).
Another less common example, if a tracking script works by calling a file with code, it can cause double page views on server logs, unless we exclude the file from the stats. Etc., etc...
