There was a fair bit of chatter in the blogs that I follow today that Smile's TextExpander is changing from an upgrade pricing model to a subscription model (details at their pricing page). It's an interesting decision, which I guess probably works out well for them since everyone that I see writing or talking about TextExpander seem to be daily users.
I say interesting, since it brings to my mind a question of conflicting values: the perceived value to the customer against the value to the developer. To generalize, I think upgrade pricing for software lets it become more valuable to the customer over time as they can amortize their initial investment while retaining the original functionality. From a developer's perspective (especially if they are active in maintaining software for platform updates, bugs etc) their software gets less value over time as their time investment increases with no additional return per customer.
To contast, subscription pricing makes software that is not under active feature development less valuable to the customer over time (granted if the software is still useful and saving time and/or bringing in money this point is arguable) as monetary investment increases, perhaps not in proportion to increased functionality. For developers the software increases in value over time for existing customers as well as from new customers.
I think this puts developers in an awkward position though, especially if they are providing a product with no ongoing costs for directly supporting the customer (for example server infrastructure for processing, storage, sync etc). There is an implied contact in that they are receiving funds monthly/annually from customers who should probably expect development toward improved or new features or support of new platforms where necessary.
To bring this back to TextExpander, whilst it can do some syncing of snippets, the way that I use it is all locally. It isn't something that requires Smile's infrastructure, and while bug fixes get supplied through updates, quite frankly the customer should never have to pay for bug fixes (compatibility updates are another matter since in most cases developers shouldn't be held responsible for platform changes unless they willfully ignore deprecate APIs etc). I hope that Smile publishes some information in the future about how this changes the fortunes of TextExpander. In the meantime I'll just continue using the old version.
It's also probably useful looking at the patronage model that something like Overcast uses, although this has changed slightly with the 2.5 update that Marco released recently. Prior to 2.5 Overcast had a patronage donation setup where you could donate for a certain amount of time, and initially it was just that: a donation. This felt different, even though it was in a very similar sort of setup to TextExpander. Both bits of software save you time, TextExpander by expanding snippets of text, and Overcast by speeding up podcasts and skipping silence. I was quite happy to give a yearly "subscription" for Overcast, partly because I use the hell out of it, but partly because I knew if I stopped, the software would still work.
(Post 2.5 Marco has added some new patron-only features to Overcast like customer file uploads and a separate visual theme, although I don't use either of those.)