I have never fully understood how this oft-mentioned “if your app shows good numbers for ad networks, they will send better ads your way” … ?
What is it - they are expecting that some app may not be showing ads right - so it is pointless to waste bandwidth delivering an ad to somewhoe who is not a doing that ad justice … ?
Sorry if there was any confusion with my previous statement.
It might be best to explain it like investing in stocks. If you have money to invest, are you going to invest in a stock like Apple or Google that has a known history of performing well or are you going to invest all your money in penny stocks where you don’t have a whole lot of previous knowledge of how the stock will perform?
Ad networks have servers and platforms with algorithms that funnel higher earning ads to apps that have performed well with them in the past on previous ad campaigns. Not always, but many times ad agencies are looking for a campaign to have an overall ctr that they are looking to accomplish. From the get go the ad ops team of a network knows for the most part which apps have previously performed well with similar campaigns in the past and will start optimizing the campaign to target the better performing sites.
Thus all things being equal, if one site has a better ctr than another it will receive the “better ad” than the other app with a lower ctr.
I could understand Admob doing this type of automated “rating” of an app essentially - esp. for the “long tail” of small apps (there would be too many to do manual assessment). I would expect Admob would be building up stats on each app to see what type of ads perform on that app - so future targeting is more efficient. If so, this WOULD lead to rising or falling eCPM depending on the app. Optimally, this would benefit the app developer as well - since admob would be making sure the ads that are shown ARE clicked (and the developer paid) - so developer would presumably benefit from an intelligent showing of the “right” ads for that app. This would inevitably lead to differing eCPM across apps. In fact EVEN if admob didn’t do any profiling - showing the SAME ads to different apps will get a different eCPM result (showing game ads in a tool app may give worse performance than showing game ads in a game app - where users are more likely to click on game ads).
For advertisers - do they have a console where they can select apps - or more interestingly do programmatic targeting of apps ? Or would this be handled completely by admob ?
With only very large marketing budgets - say those looking for 1M banner impressions on Angry Birds - those I suppose could happen manually - but otherwise it seems difficult for advertisers to be selecting apps which have 1K impressions (too much work) ?
Though there probably is a “long tail” of advertisers also - those with small budgets.
The idea that Admob would be profiling apps - makes me wonder that my practice of not using the testing ID while I am testing (and not clicking on ads) may be giving Admob the wrong idea about the app. Once it is released on Google Play it would generate many more downloads and impressions would soar. So hopefully that early period of running in house with the real Admob ID does not convince Admob that the app is pathetic. If this misconception is there then I would guess Admob maybe doing periodic reviews so eventually that would be corrected so profile represents how the app is behaving when it is public.