Ask HN: Any way to find why users uninstall my app?

With free/in-app apps, a popular workflow for selecting an app to actually use is install half a dozen of the offerings, give each a test-ride worth a few seconds and keep the one that seems to best match personal usage patterns. 20% retention after the first five minutes might actually be a really good value!

If you do not have a means to tell those short-lived uninstallations from uninstallations by long-term users, those numbers are about as meaningless as comparing the number of play store page visits to the number of installations. Remember when the first online retailers panicked about abandoned virtual shopping carts?

You should set up analytics in the app which tell you at least whivh pages they open in the app before uninstalling. Do you even know if they ever open it? Perhaps they are installing six Todo apps, they like the second one so they uninstall the others without trying them - or perhaps they look at yours and it doesn't win.

If it were me installing your app, I would uninstall it for lack of enough features. (I did not actually try your app, maybe I will.)

Some things that appear to be missing (from the google page, I didn't actually try it):

A way to sleep/snooze a reminder for a specific period of time.

Per-todo item auto-sleep rules. (Sleep, and just stay as a notification? Sleep and go away? Sleep and alarm again?)

Options for what kind of alert - noisy, quiet, speak the alert, buzz, start quiet, get louder. (All per todo, not global.)

Does it turn on the screen when there is an alarm? It shouldn't.

A guarantee that no alerts will be "lost" if the device is off when the reminder should have gone off.

Way, way more flexible options for reminders. You need things like: Remind for 4 days, then sleep for 2 weeks. Or every other week, but only on these days of the week. There are many patterns.

They way you program that is start with a basic repeat, then you add exclusions, "not on these, days/weeks/months/dayofweek".

Next you add "don't start until date", and "stop after date".

Then you add unlimited "don't alarm on this date", and "alarm on this date", where they just add a list of them.

Next you need multiple alarms per todo - the first one is just a notification, and if I dismiss it the rest of the alarms go away. The next alarm is a short quiet one. And then a loud one. But the user picks what kind of alarm, and how much before the event. They can add as many pre-alarms as they like.

I have not found a good reminder program, very very very few have flexible enough reminders. The best I've found is https://play.google.com/store/apps/details?id=com.splunchy.a... - install it just to check out how to do complex alarms, and complex repetitions. It's a beta though, with bugs, so I'm not suggesting it for use, just to see examples.

Do you want to make a comprehensive alarm program? Or do you want something simple? There is a market for both.

Each time I install an app that says

Offers in-app purchases

I end up uninstalling it quickly (after 1 try), those ads in apps are really annoying for the user.

It's not the developers fault, but the ad system on either Android or iOS are really annoying, even if the app has all the features, for me the ads kill it.

It's kind of a crowded category. If users try five different apps and then pick one, most of the apps will have very high uninstall rates.

In that field I'd say your problem isn't really about your app, it's just hard for people to get into the habit of using an app, so a high uninstall app once people give up should be expected.

One thing that I haven't seen yet (and I haven't tried yours) is a way to add a task through the notification screen. I don't want a widget on my screen, and having to flip to the right screen to load the app is enough work that I just don't do it.

Are you asking them to give them their email ids, etc to you? If yes, then why don't you email users who haven't used the app in a long time, about news features, etc and also include a feedback link there. You might get some responses.

Also, try to find the uninstall rates for other todo apps. I personally in a moment of self realization decide to make todos, but the very next day I get fed up and uninstall the app, maybe thats common across a lot of users?

App evolution is purely Darwinian. There is no $$ lock-in, so as soon as the user finds what he feels is a better one, yours is killed off. As to what is a better one, it might be that it has more functions, less crashing, easier to use, lower CPU burden, better marketing BS and so on - nothing stands still. So you have to fine tune your ap to make it evolve faster than the others into a better thing.

Do you have a definition for an active user? If so how the active vs. passive user rate?

Arrange user tests, ask people who are unfamiliar to your product to download and try. You could learn a ton from that experience watching people using your baby.

Another option is to record user actions in an anonymised way, upload that metrics and analyse, compare it to installation where they haven't removed it.

Is there a way to find for how long they use the app before deleting it? Can you track the activity of the user, e.g. number of tasks added per day? Or, in other words, are you losing long-term users or the ones that are trying out 5 different todo/reminder apps just to pick the one they like the most?

It's not free. Also an app with in-app purchases is not free. I think in a crowded category at least one app is going to be genuinely free without ads or freemium.

PS: I believe apps should be paid for.

try to see how users behave on your app by with usertesting.com or equivalent service

Please consider that in developing countries, most of the phones (Android, most of them) have precious little memory, users often have to delete apps, in order, to say, take an extra photo.