In a recent discussion about improving earnings of forums - http://forums.digitalpoint.com/showthread.php?t=266350 - it dawned on my that not many people consider their CTR per visitor instead of CTR per impression. I personally think at the end of the day improving CTR per visitor is what is most important because you can't reasonably expect one visitor if they view 100 pages of your site to click 10 ads, and even if they did it is my understanding that Google starts to discount clicks from the same user if the clicks are within too brief a period anyway. So if you have never considered it before it might be worthwhile just looking what your CTR is per visitor. How do you work it out? Just divide the number of clicks your receive from a website by the number of unique visitors in a day, this is best done using Google Analytics from my experience because it is on a similar time setting to Google AdSense. I think most people will find the results quite interesting, and at the end of the day this in my opinion is the figure you want to improve not CTR per impression.
Excellent point and I do exactly what you suggest by comparing my analytics visitors to my impressions and clicks. The visitors reported by analytics probably does not match exactly since impressions in adsense don't match pageviews in analytics but it's close enough, especially if making relative comparisons.
funny, because i have always done my best to keep my CTRv low.. This is because if my severe case of GoogleBanAphobia.. Many publishers that suffer GoogleBanAphobia use tools like adlogger to limit CTRv ..
Maybe i'm confused by your post, but the way i read it, you are suggesting that it is important to improve your CTRv (clicks per visitor).. Because of the fear of being banned by adsense (GoogleBanAphobia) for invalid clicks, or too many clicks per visitor, many publishers use Adlogger or other tools to stop displaying ads to visitors that click "too many" times over x-period of time. Preventing the very thing you suggest to improve.. A paradox.. Or maybe i'm just confused.
A paradox or maybe just different ways of looking at it. For me, I've been noticing a steady increase in the number of pageviews per visit. So, my effective click per impression would actually decrease. I don't use the click per visit so much as something I try to increase, but more of a gauge. Is it staying consistent even if visitors are viewing more pages? Or, how many visitors do I need to have to be making X? I haven't really looked at it as something increase, though within reason it wouldn't be entirely bad. You could get your click per visit up to 100% but if each visitor is viewing 10 pages, are you going to run into trouble for a high CTR? I really don't know the answer to that!
I'm not implying trying to get visitors to click multiple ads, rather try to get more visitors to click "a" ad. Basically what I'm saying is if you have 1000 visitors to a site in a day and you are getting 500 clicks on your ads that is a good CTR per visitor at 50% but the way Google reports is is CTR per impression, which means that if on average every visitors to your site views 20 pages of you site your CTR might be 2.5% which many might consider to be a poor CTR. You see what I'm saying now - I just think that too many people focus on the traditional CTR per impression rather than CTR per visitor. I highly doubt Google has any problem with CTR per visitor of 100% - as long as the clicks are being obtained via means that are within their TOS. After all Google wants you to get as many clicks as possible because clicks means $$$ to them as well.
Personally, I couldn't care less what my CTR is. I am interested in the percentage of visitors converted (clicking ads), but I really focus on revenue per visitor. That to me is the key number, particularly if your are paying to promote your website.
The CTR displayed in Adsense is click thru rate based on page views. Unless your visitors only view one page, the numbers aren't close to the same. I should say, I don't pay much attention to CTR on site-wide basis, but I do find it useful in terms of individual pages.