Finally have some real data of my own to report - mainly about AppBrain or the offerwall format:
- offerwalls
- banner ads for “More Apps”
- Admob
- Millennial Media (mmedia)
- Greystripe interstitials
- Greystripe banners
Firstly Greystripe banners are giving 0% fill rate.
Greystripe interstitials are giving same eCPM as banners - which kind of destroys the motivation for disrupting user - may as well show a banner ad. Now this maybe just Greystripe doing some initial testing. But one thing is for sure - there is very little feedback from Greystripe about exactly what is happening with your ad time/space. Some developers have reported here that Greystripe can run public service ads (for which you get no revenue) - that is great, but is at developer expense basically (with decisions made by Greystripe and not the developer regarding this matter).
To be noted is that if a developer is going to incorporate interstitial - and thereby jump through hoops himself and subject the user to such as well - then there better be greater eCPM (or overall revenue from that). Also one cannot show interstitials all the time (like banner ads can) - i.e. your total presentations of interstitials will be much less - so the eCPM better be huge c.f. banner ads for them to be equal to the revenue from banner ads even.
In my app - the total impressions of the interstitial is less than 2000 impressions over a few days period.
In comparison the Admob mediation webpage shows 20,000 network requests - impressions obviously lower than that depending on fill rate.
So you can see that banner ads get shown about 10x more often than banner ads. For the same time period, the eCPM for interstitials better be 10x if you want interstitial revenue to be say … equal to that from banner ads.
Now to be fair to Greystripe - this figure of 2000 impressions vs. 20,000 impressions from banner ads can probably be tilted to be 4000 or 8000 impressions for Greystripe interstitials - if one just shows them more often or on more occasions. So there is probably some leeway (esp. if Greystripe pays most of the time per-impression). Banner ads generally pay by click - which means their revenue may go down over time as users experience fatigue or familiarity with the type of banner ads you show. In comparison the Greystripe interstitial will continue to pay - if revenue is based on impressions.
Millennial Media banner ads are giving very very bad eCPMs - now this could be that my app is not approved yet “really” or they still in evaluation stage or whatever.
Millennial interstitials - these I evaluated before deploying my app - and I was hoping to add this as a secondary to the Greystripe interstitials. However I was not able to get their ads to show in my app (and I had the distinct feeling it might make my app more unstable - requiring more tests - so I postponed adding that and released the app anyway). In addition with Millennial Media demo app - I could see that their interstitials were very slow loading - it is basically as if a video player is loading - this delay is costly and boring to the user - and I did not get a good sense of comfort with using such a format.
I split Admob mediation banner ad delivery to same percentage between Admob banners - AppBrain banners - Greystripe banners - Millennial Media banners.
Given they are given the same opportunity, regardless of what their fill rate was or CTR etc. etc. - the end result was that:
- Admob was giving some revenue - but mmedia was giving a fraction of that (as said above - maybe it is that they are still “evaluating” the app)
- Greystripe banner ads were not being delivered - 0% fill rate - maybe Greystripe has no banner inventory.
- AppBrain banner ads were actually generating revenue - this was the most surprising part of these stats (for me)
As some may recall - I have posed the question here to others - including gordonb etc. - about what the utility would be of presenting AppBrain banner ads - when all they are are “More Apps” - plus if you are showing the AppBrain offerwall every 3 days on exit from your app (as they recommend) - what is the new thing you add by showing banner ads that just say “More Apps” (and would not just putting a “More Apps” link on your screen be simpler).
Now I have not tested the putting of “More Apps” link on the main screen - say a small icon on bottom-left etc. - I suspect it MAY in fact be simialr to a banner ad (though banner ads are dynamic so may catch the user’s eye a bit more and get better click rate etc.).
However, one thing was clear from the stats - the AppBrain banner ads were leading to nearly the same or more revenue than the offerwall I was showing “once per 3 days” at the end of the app.
On the other hand only 3 or so days have elapsed since app was released - so maybe these stats have not included a couple of the 3 day cycles and maybe misleading.
But one thing is clear - the AppBrain banner ads do deliver.
A caveat to consider however is that - as discussed by others here previously - new users or a new app CAN tend to get higher number of clicks from users - as they are new to the app and click all over the place - over time they get familiar and probably avoid clicking things they know or have seen previously …
This would suggest the high AppBrain banner and offerwall click rates may be higher now and should fall later.
But it WOULD be interesting if it turns out in the LONG RUN as well IF the AppBrain banner ads give same revenue as the offerwall at the end (shown once per 3 days).
Another thing to consider is that - once one has this sense that AppBrain banner ads DO give revenue - the next question to ask is SHOULD one give more ad traffic (or impressions) to the AppBrain banner ? For example by shifting away allocation from Millennial Media banner ads to AppBrain banner ads ?
Intuitively one can sense that this may not be such a linear relationship - that is, there is probably a limit or saturation point - to how many times you can show a “More Apps” banner - i.e. the user will get tired.
And that perhaps the novelty of “More Apps” may lie in them being alternated with other content - that is, AppBrain banner ads MAY work better if they are mixed in with other banner ads - just saying this - one would need to run an experiment to confirm this sense.
So these are open questions …
Perhaps others here have some sense of - I think gordonb or someone mentioned that when they presented the AppBrain banner ads or offer wall more time they got more revenue.
If so that would suggest that one should experiment with showing this type of stuff more often until one senses it does not help revenue - then one can back off and mix it with some other banner ads etc. - to give the banner ad space some “novelty”.
Anyway … some thoughts …
I have given the majority of the allocation for Millennial Media banner ads to AppBrain banner ads - to see if that increases the revenue linearly for AppBrain banner ads. That is as presented there is STILL some leeway for showing them MORE often … before one hits the saturation point mentioned above (where user has seen “More Apps” so often they tune them out).
[hr]
Revisiting some of the comments by gordonb on Appwalls …
-
i think your display strategy is good, works perfectly for the type of app and that is key. you should also show it on app exit though at the minimum, and possibly at app launch. you will see that this will greatly increase your revenue.
-
i like greystripe a lot, they are great for rich media however the CPM’s do not come close to Airpush SmartWall and their fill rate is weak because there aren’t many Rich Media advertisers out there yet.
-
i don’t get negative ratings due to SmartWall…push and icon cause a bit, but not SmartWall.
I have also begun thinking along the lines of how best to maximize the Appwall format - as that seems to be ONE place where there is (at least at the moment) money flowing in - as the commercial app producers try to buy their way into greater installs to get into Google’s top 25 lists (from which point they have a self-sustaining model as hordes of users start to see their app).
So as gordonb suggests one way is to show the AppWall unashamedly at startup even (just does not seem like a good way - and will spoil the app’s look and feel) - but if one is starving for revenue - then it maybe worth it (for the developer) - in lieu of a minor inconvenience to the end-user.
In any case, that is something one should build into their app (perhaps switchable) - and one can then test it to see how it does for a couple of days.
The only thing which intuitively (i.e. without data but just one’s own sense) one senses is that Appwall here and there all over the place may ruin their novelty factor - and user may tune them out. AppBrain suggests showing it once every 3 days at app exit - as something that seems to work sustainably … and that makes sense.
Of course the developer will want to test the boundaries themselves as doubling the presentation rate of the Appwall may give short-term doubling of revenue … question is if that is sustainable and can it lead to a decline in long-term value of that Appwall - or is it that DESPITE that tune-out - the greater presentation still leads to if not double then a 50% improvement. Probably as one doubles or triples presentation of Appwalls - there maybe less yield in revenue - but maybe it still goes up - so a developer may think they can just keep increasing it’s presentation frequency - however the payback may be less and less (and at great cost to user experience).
The other thing I was considering was to replace the Greystripe interstitial (which I present on occasion - as a “screensaver” type of thing.
Now I was considering - what if I show the Appwall instead - as the screensave type etc.
Will users get pissed off at seeing the Appwall - with Greystripe interstitial ads - at leats they would not have one manufacturer to be pissed off at - they would vary by ad - and they may like some ads.
What would happen to Appwalls.
Or is it that users see some value - maybe AppBrain presents good app choices which really ARE good - or since these are paid promotions - usually there is money behind these apps so they have a certain minimum level of professionalism - so could it be that there is VALUE to the presentation of the AppBrain (or similar other high-quality Appwall presenter) ?
If so - it COULD be that greater presentation may not piss off user as much - if they have a sense that they usually DO find something interesting in the “More Apps” thing.
All this perhaps need experimentation - i.e. double the presentation of the AppBrain banner ad - and see if that doubles the revenue from that - that would confirm that are still FAR from the saturation point.
Similarly replace the Greystripe interstitial ads with the AppBrain appwall - see if that reduces the revenue yield from the AppBrain banner ads (i.e. would suggest users are getting tired of the whole “More Apps” thing because you are showing it so often etc.).
Anyone done some experiments like this …
