This. With very few exceptions (content subscriptions (e.g. Netflix) and services that have a server-side component core to its functionality (e.g. Dropbox)), I avoid subscriptions like the plague.

With applications, I don't choose between more or less expensive subscription, but I'll chose your competitor who sells upfront.

The way I see it: unless you’re putting in a bunch of monthly resources for me (in terms of content creation or server resources), you have no business charging me a recurring fee. If it’s essentially an offline app (or easily could be) then I’m either paying up front or not at all. I’ve passed up on many apps that I would have happily paid for (in some cases up to about $100), but was unwilling to pay a subscription of any cost.

ES Explorer, I'm looking at you

What are your feelings on applications which have a server-side component which isn't necessary to it's functioning? For example JetBrains products with their plugin store or Unity with the asset store?

Note that JetBrains products come with a perpetual fallback license, so you can forever use the version of the software available at the time you bought it, just without any updates. https://sales.jetbrains.com/hc/en-gb/articles/207240845-What...

Easy upgrades are already a thing, though... just gate new functionality behind IAP, at a low cost per set of related features. Maybe toss in any general quality-of-life improvements as part of the free functionality, too. Call it the "Paradox DLC" model.

There is other side to the story. If your app requires a backend server, it is financially prudent to pay a recurring fee that way hosting charges are taken care of and you can be somewhat sure the service will last as long as there are enough users.

There are plenty of games and apps that are dead because the developers no longer make new sales and the previous money is all spent and the business become unviable and they shutdown the servers.

Recurring charges gives the incentive for the developer to fix bugs and maintain the codebase and not abandon it completely once they loose interest, if original developer looses internest recurring revenue makes the code/service valuable so that a new developer can buy who makes enough in maintaining it.

If the app/service you bought does not have enough paying users to even pay for hosting servers or a part time developer time to maintain it, then no revenue model will save it anyway.

Higher upfront revenue for the onetime purchase option doesn't necessarily correspond to higher lifetime revenue per visitor.

Upgrades are still a good option but I didn't discuss them in the article because they aren't supported on the App Store.

One caveat with upgrades is it's difficult to maintain many version of the same app. Now with frequent hardware and OS upgrades (yearly!) it's even more expensive than ever to maintain many versions.

What does Apple need to understand? I don't believe there are any rules about publishing MyApp2.0 and then heavy advertising it in your previous versions.

One pain is sandboxing makes it difficult to migrate data from v1 to v2. You need to build an elaborately kludgey data export/import feature by hand, in both versions.

More importantly, how do you charge upgrade pricing? Any user who encounters 2.0 on the store is charged the same price, whether they own 1.0 or not. There is no easy way to give 1.0 owners a discounted price.

There are elaborate kludges to work around these limitations with a great deal of work. Apple could make producing upgrade versions easy, like they’ve done with subscriptions, so devs can focus on improving their product value instead of purchasing mechanisms.