I’ve thought about similar issues (since I cannot sell Google Payment related stuff from my location - I am talking about my thoughts about which features to offer for IAP via GetJar/Tapjoy).
My general impression is that users don’t care as much about ad removal as developers would like to think (or I thought).
Banner ads are least bothersome for users - though interstitial ads if presented in sufficient nuisance frequency would be major issue for users (but then they would have downrated your app from that already).
The reality is that you would optimally be showing interstitials at almost the border of being a pain (which would risk bad ratings) and being not bothering (which would risk ad removal being not high priority for users). Removal of such a presentation of ads therefore falls very close to being not something high on users’ agenda.
My feeling is that (and above sentence is one logical reasoning for why this maybe true) is that ad removal is not terribly high on user’s agenda (for a “nice” app which is already presenting ads not very often).
This leaves most OTHER features as being the main thing to be removed. So for IAP to work, those features should be additional (so users don’t badmouth your default app) - so additional bonus type stuff which users feel are great.
AND for “completeness” you could tack on the “removal of ads” to that also.
Now this naturally raises the question of “why must I lose ad revenue if users are not bothered by it ANYWAY” (by above reasoning). And the way I reasoned through this is that - the NUMBERS for IAP are SO LOW generally that they will be a FRACTION of your total users.
So you may see (say) IAP giving 20% of total revenue you earn - maybe higher for others.
However, the NUMBER of users actually responsible for that purchase (and this is similar to number of users who actually clicked through your ads etc.) - is MINISCULE compared to your total DAU (Daily Active Users).
So if you make $10 daily off an app from IAP - while your revenue from ads is $40 - then that IAP users are just 10 or so probably for that day ! And your DAU maybe 5000 users !
So VERY CLEARLY you lose ALMOST NOTHING by giving “remove ads as a bonus” with any IAP purchase to the user ! I think this logic should be quite compelling. Basically the number of IAP users are actually SO SMALL compared to your DAU that you LOSE NOTHING by giving “remove ads” to those IAP buyers. So this means you can “sweeten the deal” by tacking on “remove ads” to your other purchases. Or in other words ONE USER is SO USELESS for ad revenue that you can easily exempt a single user without worrying too much about revenue loss.
Now an aside on the IRONY of ad revenue sources !!
It is an irony of modern app development (or android app/user behavior/preferences) that the MAJORITY of revenue you are making (esp. from offerwalls and “do you want another app”) are essentially NON-LOYAL users of your app ! I have not done a study on whether the clicking users are actually returning users or new users (actually I DID do something close to that - with tiered Leadbolt HTML AppWall URLs - where I happened to pick DIFFERENT URLs to fetch depending on “how many days since app first installed” - and the first days was clearly the highest revenue - while later days DID on some occasion wind up being disproportionately higher in terms of revenue/numbers i.e. eCPM etc. - the second, third day returning user numbers fall off so rapidly in absolute terms that you COULD say that majority of revenue you are making IS from basically first day users).
And since from first day users you maybe losing out on 80% of users (who uninstall immediately or soon after) - maybe less for a great app - but this is the irony of it - that for most apps (and esp. so-so apps) you are making money on “not interested in your app” users !!
And this is PRECISELY WHY a lot of bad apps, spam apps etc. ARE able to make money from that. Because if the user doesn’t like the app, they are JUST AS LIKELY to click on “want to see another app” (or maybe more !) than a user who actually LIKES your app !
This therefore establishes an UPPER BOUND on how valuable an interested user actually maybe for your app (in terms of ad revenue). If the users who are NOT interested in your app are MORE valuable - then clearly the users interested in your app (and those who have gotten to the stage of IAP purchase etc.) are NOT any more valuable than the average DAU user of your app (and as stated above perhapd maybe even LESS valuable in that they maybe less bothered about clicking on ads if they actually like your app ?).
Here I am assuming an interested user is only using your app once - and this is a first approximation - because typically over a period of a week they may still only use it once. But over a lifetime, such users WOULD be spending more time in your app and thus it COULD be argued that interested users are more valuable (per user) than a non-interested user because interested user has 10x the exposure to your app (they may use your app 10 times over a lifetime).
However, it remains to be seen - i.e. explicit empirical data - HOW much valuable a non-interested user (using your app just once) is vs a typical interested user who uses your app 10 times - is that interested users likely to be 10x as valuable or is he not as valuable as one might naively think (i.e. 10x) because an interested user maybe tuning out ads and may not be that type of user (i.e. maybe a “one-app person” i.e. monogamous etc. !). It would be interesting to see that.
The question to ask is that if you have 5000 DAU (Daily Active Users) - and your IAP purchasers usually number in 5-10 users - then are those 5-10 users responsible for a significant chunk of your ad revenue ? Or is it all those other users who just use your app once.
My own feeling is that these 5-10 users could SAFELY be offered “remove ads also” without perhaps any real affect on your ad revenue. But as stated above … this is a handwaving argument and maybe true only to first approximation. There maybe some very odd statistical skew (like “hey it turns out ALL the revenue you earn is actually from all the allegiance-to-your-app users so by making ads free you are losing out on long term revenue”) - but it is probably a second degree effect which would need empirical data to be shockingly convincing.