early this morning the CTR is already at nearly 5%, so it doubled, the eCPM doubled as well
Ok, but has the total revenue gone up or down ?
I cannot tell you exactly but as I see, currently it is going down. But that was expected in the beginning.
I hope that the better CTR will lead to higher-CPC ads which will than hopefully increase the revenue.
This is also a little bit of experiment and most of what is currently happen is more because of the very high ad-refresh rate of 10 minutes.
If it does not work out, I can switch back to a higher refresh rate as I like, but for now it is experimental xD
I got the numbers from yesterday. The CTR was at 4% and the overall revenue was 75$ - that is only 5$ less than the day before at with nearly half of the impressions. 2100 clicks instead of 2700 the day before. eCPM was 1.27 instead of 0.71 the day before
Quite interesting
Looks like it is working how you hoped.
If this higher CTR starts giving you better ads with even better CPC this could be very good.
That’s the big question - theoretically a higher CTR should lead to higher CPC ads (if available) … that is logical and also the different ad networks does also suggest to not waste impressions.
Today I see the current uncorrected results which are very promising. Normally, this values will be better on the next day, when the stats get updated.
The current CTR for today is still at 4.71 and the eCPM at 1.35. Impressions are much less than usual. So that is already a gain to yesterday and can only get better.
But at the end, revenue is, what counts and revenue is lower than usual - so it really depends on the CPC now - the big unknown: will it get better? xD
Next thing to do is getting rid of adsense backfill, look at the fillrate and make a try on admob mediation
day not fully over, but not much more will change - impressing!
CTR of 6%, eCPM of 2$ with only a 1/3 of the impressions. Curious if I got any adsense impressions at all and if the stats get some correction tomorrow; will report again 
Unfortunately, I cannot tell if the close button have any influence…
What would happen if you removed the “Close Ad” button, and tell the user that if you click the Ad, it will disappear for X minutes? This way, you get a bonus ad click out of it.
I’m assuming this is even possible via some sort of onclicklistener for the ad you are using.
This is against the rules for ad-mob, and every advertiser I know of. You might get away with it for a while, but if ad-mob found out they could cancel your account.
Yes, it is against the ToS - and also not so easy to do, there is no click event for it, you need to do it with tricks.
The Close Button itself should be no problem … it is like interacting directly with the ad … I believe and also youtube does it the same way (which is google)
I also thought about a refresh button … in case someone wants to see another ad … 
That was the first time ever, the revenue was DOWN corrected xD - after the regular correction the revenue yesterday was 50ct less than before xD
So it seems, that there were no adsense ads displayed any more - therefore, it stays the same as I reported above.
The average CPC (Cost per Click) has not changed very much currently:
0.030 on monday (that was regular)
0.035 on tuesday
0.034 on wednesday
0.035 on thursday
0.031 on friday
I’ve updated the CPC numbers in the post above. CTR at 6% but slightly raising over the last days
Still… that is 3-4 cents per click. Two weeks ago, I was getting 40 cents … but now I am down to 4 cents again 
I wish I knew how to get more per click!
40 ct per click would be very nice
currently the CTR is still raising ~0.4% every day while the average CPC does not change very much.
Because I do not get any stat updates at the next day any more, I am pretty sure, that I do not need to get backfilled with adsense.
So, when the CPC is not raising it seems confirmed, that a 10 minute refresh is a bit too long because the CTR (or the CPC) cannot compensate the lesser impressions.
I have put up a small article about my Experiment - enjoy reading
http://www.reiti.net/personal-blog/blog/103/publishers-ctr-vs-impressions
Hi,
After reading this I have analyzed a bit stats from my own app (Solitaire Champion). The reason why it is meaningful to this experiment is that I’ve made a change in the app few weeks ago. Before I had the ad banner displayed quite close to the touchable game area which most likely was resulting in some accidental clicks. I decided to do the right thing and not risk being banned by Google for fraud clicks and moved it a bit further away. Guess I was right - CTR dropped 3x from around 1.6% to 0.5%. I hurt but hey, it means I was ‘stealing’ two of every three clicks. Now - when I just calculated CPC for before and after it is pretty much the same. In fact now it is a bit better (3.5c vs 3.1c) but that may be just a statistical noise.
So my case confirms what your observations seems to suggest - there’s no direct correlation between CTR and CPC.
Max
At least not with admob - we also have to take into account, that on their frontpage (website) they boast about how many ads they delivered. That left me with the thinking, that they do aim for maximum impressions instead of maximum CTR.
Another clue is the maximum of 120 seconds you can configure on their website for refresh rate. Maybe they do want you to display as much ads as possible xD
But thanks a lot for sharing your experiences. I could still switch back to 120 sec whenever and also I would be curious what happens, when I use another ad network also, where they do look on the CTR
Also, I still do not know if admob does refresh automatically once the ad was clicked … I never clicked my ads, so I just don’t know - does anybody know that?
I’ve just found out, that I already have a game with a very very high CTR out in market
Shape Masher is such a candidate, is generates merely 30 Impressions per day (no refresh) and 0-5 clicks.
This game has a CTR of up to 20% so I decided to calculate the CPC out of it with interesting results. Most of the time (not always) the CPC was over 0.05 - it had reached also 0.09 one time.
Because most of the time the CPC is higher than usual I don’t think that this is random BUT I do think that admob just lacks a decent amount of such high CPC ads.
If it would be possible to get a CPC of 0.09 all the time in combination with a CTR of (currently) 7% with (currently) 35k impressions (is nearly 2.5k clicks) that would more than tripple my revenue (as it was before the experiment). And that would be WOW because that would be nearly the revenue of the peak times of Robo Miner (when it was 2 million imps/day)
But I don’t think currently, that admob can serve that much high CPC ads and that’s why my experiment does not pay off. That problem should be solved when I integrate another ad network and do some mediaton - unfortunately I just started with other networks, so I do not know which network to use … Jumptap seems to work good in Mopub, so maybe I will try that
update with some changes and a conclusion
http://www.reiti.net/personal-blog/blog/105/ctr-vs-impressions-conclusion
Actually, you want to plot the income vs refresh rate. Then aim for the sweet spot! For WordHero, this is around 65 seconds.