i don't find it too compelling, but this test could potentially help you understand, to some degree, how important battery consumption of an app actually is to users.
so you run a test where you intentionally consume more of some cohort of users' battery power (you could have several cohorts where you consume more and more power, even). and you look to see if the rate at which your app is deleted/force-quit by the cohorts with more power usage goes up. if it does then you can assume that you're overdoing the battery drain. if it doesn't then you can assume users don't really care (or don't care enough; maybe they're unhappy but your app is too important to them to delete).
why this could be useful is when you're deciding what to prioritize - if you've got data saying that users don't care about excessive battery consumption and they'll keep using the app anyway, you can argue against optimizing for battery life in future development, presumably letting your developers do things faster/more lazily. or, it could show that battery life is super important and be a valuable argument to prioritize power optimization work in the name of keeping your users from jumping ship.
personally i'd rather just presume that battery life is important and that optimizing for efficient use of our users' batteries is the right thing to do, regardless of hard data, but i'm sure there are people out there that think differently.
hard to make that study work, it presupposes that users (a) understand that their battery life is decreasing and (b) understand that this application is responsible for it (and to what degree). Those are two big if's, it's more likely they'll chalk it up to battery aging than to a malicious application that used to be well-behaved. That doesn't however prove that users don't care - it only takes one article with a headline "Phone draining quickly? Facebook battery usage has increased by XX% the past year and responsible for majority of its users battery drain" to completely swing the pendulum to the other side where more users are deleting your app because of this new reputation, than would proportional to the actual battery drain; but without that trigger the study is not complete.
so you run a test where you intentionally consume more of some cohort of users' battery power (you could have several cohorts where you consume more and more power, even). and you look to see if the rate at which your app is deleted/force-quit by the cohorts with more power usage goes up. if it does then you can assume that you're overdoing the battery drain. if it doesn't then you can assume users don't really care (or don't care enough; maybe they're unhappy but your app is too important to them to delete).
why this could be useful is when you're deciding what to prioritize - if you've got data saying that users don't care about excessive battery consumption and they'll keep using the app anyway, you can argue against optimizing for battery life in future development, presumably letting your developers do things faster/more lazily. or, it could show that battery life is super important and be a valuable argument to prioritize power optimization work in the name of keeping your users from jumping ship.
personally i'd rather just presume that battery life is important and that optimizing for efficient use of our users' batteries is the right thing to do, regardless of hard data, but i'm sure there are people out there that think differently.