Forum Moderators: martinibuster
The layout on this particular site is quite simple, just like Google's search results (used to be) there are one to three skyscrapers of text only ads down one side of the page, and I've discovered that Ad Units seem to work well in locations where you expect visitors may exit the site. An example being the bottom of the page where visitors expect site wide navigation. This Ad Units are well "displaced" from site navigation itself. For this site, I've always seen earnings decline with fewer ad units. There is a lot of content and many diverse mostly targeted ads are displayed, except when there are quirks as reported in several other posts (weight loss, etc.)
I also believe at this time I have all "new code", which for some appears to be an issue.
Has anyone else seen this problem or are you aware of it? It's interesting when I ask for Ad Unit results, of course, the CTR is through the roof. I have clicks with no impressions, but everything does make sense in the combined report and Link Unit report. I would report this to Adsense but lately they have been unresponsive and there is the concern of over-reaction on their part.
See this post:
Email Support useless? [webmasterworld.com]
In general, it appears that if you have both link units and ad units on the same page, the "page impressions" are divided between the two when you view the separate "ad units" and "link units" reports for that channel. In other words, the sum of page impressions in the two reports equals the page impressions in the default/combined report. Of course this is misleading, because, as you say, the CTRs on the two individual reports are inaccurately high. In reality, the page impressions should be equal in both reports if you're viewing, for example, a URL channel for a single page that contains both an ad unit and a link unit.
I have to assume Google is aware of the issue since it's the same for everyone I've talked to. The workarounds seem to be either a) disregard the impressions in the individual reports, b) rely only on the "combined" report, or c) use separate custom channels for ad units and link units.
Thanks for posting. Early on in the Adsense release, just about every new design feature had some type of statistics bug. I know if I look back historically in the data for 2003 and 2004, I have to take into account the fact that page impressions may be double or triple actual values.
So I'm actually still a little suspicious that Google hasn't actually figured this quirk out yet. I don't think they use their own product that much.
I didn't have any inconsistencies with multiple Ad units and one Link unit, but since adding the 2nd & 3rd link units, this brought the problem to light.
I see I mistyped in my original post, I meant to say link units work well where you expect visitors to exit the site.
Another note, I was mixing "old" and "new" code. I've now "upgraded" to all new code which made no difference in the statistics quirks.
