What was the download rate you were getting prior to the AppBrain injection (i.e. this being related to your “keyword searchability” of description/title etc.
So you went from X downloads per day to 1000 downloads per day - with rise on Google Play category list to 200.
Breaking that down somewhat …
Prior to AppBrain campaign - getting X downloads per day - this one could attribute to the “keyword searchability” of your description/title. Left on it’s own this may propel the app upwards as well - did you leave your app unpromoted to have a sense what the natural ranking progression was - or had it reached a point of stagnation at X downloads per day a few days after launch ?
So as soon as you started AppBrain campaign - the downloads immediately went to 1000 downloads per day - this would then be DIRECTLY as a result of AppBrain (or 1000 minus X download per day to be precise).
Thanks to the increased downloads, your rankings would start to rise - leading to a rise in the X downloads to Y downloads per day.
Eventually you would be getting 1000 downloads from AppBrain + Y downloads per day.
On the last day of your AppBrain campaign were you able to measure the downloads attributable to your 200 ranking in your category ?
After you stopped AppBrain campaign - presumably either because you ran out of allocated budget OR saw that the Y downloads were significant and so may not require AppBrain download to keep rank high.
Starting at a 1000 + Y downloads per day - which was yielding you 200 ranking in your category, you then removed AppBrain campaign - which immediately removed the 1000 downloads per day - leading to a shortfall.
This would then trigger the decline in total downloads - first dropping you in rankings by 1 or two places. But next day the downloads be even less (because lower in rankings) - thus leading to a vicious circle which eventually drops the rankings to a level where AGAIN the “keyword searchability” random-discovery-related downloads start to dominate.
The question arises whether your FINAL “keyword searchability” related downloads are bigger or the word-of-mouth related downloads.
And this I guess is the difference between the “viral” apps which keep rising - as at some point the word-of-mouth kicks in OR (more significantly even) if Google starts featuring the app etc.
Or that the app enters the top 10 rankings - so that it CONTINUES to be visible to users - I suspect that being at 200 ranking may get downloads from that (but 200 is a LONG ways down in the list). It is possible that a stable 200 ranking is reached and goes up and down PURELY on the basis of discoverability from keyword searches ? And only when the app has reached top 10-20 is where the ranking-related downloads kick in or what ? Though that can’t be the only explanation, since after all rise in rankings by 10 places (even if it goes from 200 to 190 in rankings) DOES affect total downloads - so what is happening ? A “long tail” of users are looking at the apps at 200 position in Entertainment category etc. and this is what we see impact on downloads ?
If so, getting an app to top 10-20 would have HUGE HUGE impact - and perhaps is the holy grail for new apps which spend cash to get there. When they have spent $1-5M on a new game - they can surely spend $1M in promotion (and plus they have experience from real-world promotion of games etc. to not be fazed by the expenses incurred in such promotion on Google Play).
Anyway … just rambling here - but how were your downloads doing prior to the AppBrain campaigin - i.e. did you feel your trajectory was not going up fast enough etc. or did you plan in advance (knowing what happens generally at launch) that you would use the AppBrain campaign etc.
What ranking is the app now (at 900 installs per day) ?
Thanks.
Comments for new folks here:
Other comments corroborate your preference for AppBrain - i.e. folks saying they were able to drive good download numbers with AppBrain at $0.20 per installs or similar.
While their results with Admob and other were varied - i.e. not able to drive their daily app install budget.
Do you have any guess on the “stick rate” i.e. retention for the AppBrain-derived users ? For non-AppBrain aware users - it is an AppWall - and thus is “non-incentivized” so users are shown apps and they can choose which app they like - thus retention would be higher than GetJar/Tapjoy (which are “incentivized” i.e. user could give a damn what the app is, they are just looking for the coins earned and may uninstall the app soon after). When the user CHOOSES the app to download from among a list of apps, they are probably MORE LIKELY to think it is “their app” i.e. they have selected it, and perhaps more likely to examine it in detail before they uninstall it.