I have been running an experiment around Google AdSense for the past few weeks, based on the recommendations of RN, and it looks like that the change has had a positive effect on the CPM of the ad unit where that change was applied.
The experiment goes something like this.
I have a 300×250 ad unit in the right sidebar of this site that’s placed above the fold (ATF) and it serves both image (rich media) and text ads. The ad unit is enabled for placement targeting so an advertiser can easily target it for display campaigns through Google AdWords.
Now let’s assume that the average CPM for this ad unit, according to Google AdSense reports, is $5.
To get higher CPM ads, I created a simple rule that blocked Google ads from showing up unless their CPM was, say, $6 or about 20% higher than the current average value. If the CPM was lower, an in-house ad (promoting one of my own articles from the archives) was displayed to the visitor instead of the AdSense ad.
This experiment has been live for about 8 weeks now and I think it did help in pushing-up the CPMs for that unit. Why? I don’t have an exact answer but probably because the Google ad now has to compete with an external ad in order to show up on the site and thus it raises its own CPM to outbid your in-house ad.
The other advantage is that instead of serving low-paying display ads, I can use that space to promote my older content to new visitors who may have otherwise missed those stories.
If you would like to try something similar on your own blog, read the Google DFP tutorial to understand how you may serve in-house ads alongside Google AdSense. Set a CPM threshold that is 10-20% higher than your current average and run the experiment for at least 10-15 days. Good luck!
Related: Google AdSense Optimization Tips [Presentation]