Should I be keeping track of in-app purchase subscriptions by myself?

You must use IAP if required to by the guidelines.

There are various 'packages' and 'code examples' out there for IAP.

With the exception of verifying receipts, it's not that hard.

If you have a server, even verifying receipts is not that hard.

And you will discover that non-renewing subscriptions are much easier to implement than autorenewable subscriptions. The question is whether you want that autorenew feature.

Finally, 30% is only the other side of 70%. 70% of X is larger than 100% of X/2.

Hi, thank you for the response.

The hardest thing seems to be all the factors that go into keeping the subscription records up-to-date. I am verifying the reciepts from the server, so it's not that big of a deal. But then consolidating that with Google subscriptions and possibly web is going to make it more complicated as well.

I definitely do want the autorenew, so I guess I'll just have to go ahead and do it.

I understand that, but I need to generate at least $3 a month just to come out even with the server costs from a handful of users. If I make the subscription $4.99/month it should cover it if I can get even one subscriber...the problem is that there's a lot of uncertainty. Right now I only have 3 daily active users, but the premium version is coming alongside a big update as well, so people may or may not be more willing to use the app. Oh well, I suppose at the end of the day I can just raise prices if I have to.

Also, were you referencing something specific with "X/2"?

It may be simpler than you think. Divide the task into two streams:

1) On each device, rely on updatedTransactions to set the expires date based on the transaction.transactionDate and the transaction.payment.productIdentifier properties.

2) Do whatever you must to verify the receipt and accept that value.

Everything else is extra fluff. Even #2 above.

Would it be better to just use server-to-server notifications?

Based on the number of problems reported in the In App Forum....absa****inglutely not!

Ok, I have one more question if you don't mind. Do you think it's worth it to use something like RevenueCat?

Based on the number of problems reported in the In App Forum....absa****inglutely not!

If that were the basis for making decisions, then we should not:

notarize software,

use dylibs,

submit apps,

purchase apple developer accounts.

Cute.

I was writing that there are alternatives to using server-to-server notifications to maintain and verify an autorenewable subscription. Use those alternatives. Don't chose server-to-server notifications because there are more problems reported for them than for those other methods. It is unclear if there are similar alternatives with fewer problems to:

notarize software,

use dylibs,

submit apps,

purchase apple developer accounts.

I am not familar with Revenue Cat. And I generally avoid using such services/code, I prefer coding it myself.