Forum Moderators: martinibuster
If I believe Adsense figures then I've got a nice healthy CTR that matches the normal single-figure kind of numbers you might expect for usual ad click-through rates. If I believe my own figures then I might be some way below that.
Once I account for
- bot visits,
- things my stat analyzer thinks are pages but aren't
- alternate ads (you can't coun't PSAs-- you won't know if you're serving "virtually no PSAs" unless you have an alternate defined),
- pages without adsense,
the numbers tote up quite nicely, and hte deiscrepancy due to non-JS browsers is likely to account for the difference, but I see no return on the effort required to account for it at that level.
Another forum member uses a javascript for tracking, this would be the ultimate way to go because it neatly cancels around all the user agent issues. Try that for a week and see if you don't think the numbers agree.
Overall, tracking impressions and clickthroughs is such a straightforward task that I have no doubt Google's numbers are accurate. I would even call it trivial, except that they must have to combine stats from multiple servers.