Does app size matter for install rates? I suspect not on high end mobile but this accident should give some data towards this.
Will throw a spanner into the 'we can't release that feature as it will increase app size' thinking that I've witnessed in my mobile dev career. I've never seen it actually impact metrics that much.
I'm not sure it affects install rates - but I know when I've been cleaning up my phone the apps that seem to occupy orders of magnitude more than their function would suggest, get culled first.
Makes me feel old. I'm sure 20 years ago I could have told you say what app/game had been installed from floppies and which was "CD ROM only"
My phone apps? Couldn't even guess as to their storage requirements.
Probably a good thing that storage is so plentiful we normally don't care - but cynic in me thinks that a company selling storage with a massive markup, maybe doesn't have aligned interests.
If 0.1% of people install an app, and it consumes a gig rather than 500Meg - and we sell millions of phones. That does turn into a real money pretty fast.
There are A LOT of people who ran out of space on their phones and can't install anything anymore. So if your app is 500MB, they'll first have to find 500MB worth of stuff to delete.
If they really want your app, they'll look through the storage settings on their phone, and start deleting their biggest apps.
So if your app is too big, many people won't even install it, and if they do, it'll be the first app to go if they need more space.
I just moved from android to apple and the IOS apps are 2x - 5X larger
It's really a disgusting tax and I install a lot less apps as a result.
They could easily fix this by adding intelligence in the App store, to serve up the binary for my device's architecture, but no, I have to get the universal bloated binary pig.
Example: You're about to pay your bill at a restaurant. The server tells you that they have an app that will get you a discount/loyalty points. Your cell has a 4G connection. On android, the download size is 20-50MB. No problem. On IOS, the download size is 40-240MB. Problem!
Yes it matters. My current scenario right now is our internet is down. The fastest way to reach ISP support is through Twitter DM. Twitter is not an option since their app is way over 300MB. So I download Tweetbot for only ~20MB.
I am also an Android dev and maintaining APK size has always been part of the dev process from the start. Apple is just so bad in policing their IPA size. And it is getting worse.
The problem isn't the speed, it's the ridiculous bandwidth caps imposed by ISPs. Here in Canada a fairly simple 10GB data plan runs about $75/month, if a restaurant is offering me a discount and requires their 250MB app to get it, that discount has effectively cost me $2.
Yeah, I haven't considered data caps, as here in Hungary, unlimited data is around 27eur/mo, but even capped is not that expensive (just an example: 10GB is 10eur, 50gb is 20 eur)
I have a really hard time calling those plans "unlimited" when it becomes unusably slow after your full-speed allotment. I hope the waiter told you about their app discount before you ordered, because best-case scenario that 250MB app is going to take a little over an hour to download at the 512kbps they allow.
There's a 200 MB size threshold where users on cellular connections get prompted if they're sure they want to download it, so if this pushes an app over 200 MB I would assume yes some users will bounce off of that.
And in some places 200 MB can still be a big deal, the whole world isn't on 20 GB data plans.
200MB for a 3d game with a bunch of assests... sure... but somehow there are more and more >200MB apps, that are basically a packed webapp that should be below 10MB.
It used to matter in the past, but it seems like these days the effect is not very significant...
Last year we ran a test which involved artificially bloating one of our IPAs from 160mb to 260mb* and compared the app performance metrics* before and after the change, we saw no meaningful difference between the two versions.
By using the app thinning mechanism, you can create a different IPA sizes for different devices.
High end or low end it doesn't matter. Poorly engineered is poorly engineered regardless. High end phones with lots of storage can (for now) get away with running apps that take up 2x more space than they need to. For an analogous problem try running poorly coded apps on old hardware, looking at you MS teams or zoom.
Well hold on. That seems to be begging the question. If a whole feature can be replaced by a web view (and the site is already there), is an app actually required?
Larger apps lead to more cycling behavior. Remove X and Y, install Z, then later remove Z to install something else. Biggest app that doesn't need to be installed at the moment gets removed first.
I guess the main problem is that it's most often a boiling frog problem, not a drastic change like in this case. Users may put up with long install times because they've slowly getting used to it, and usually they also can't judge whether the install size is actually justified for the features the app offers (and I have the impression that most devs also don't have much of a clue or even care).
Will throw a spanner into the 'we can't release that feature as it will increase app size' thinking that I've witnessed in my mobile dev career. I've never seen it actually impact metrics that much.