This is exactly why I was interested to hear what everyone thought.
To me, GaaS is a term that applies to the modern way of approaching this. It's something that's been refined to a specific point, even if previous games over the years could also fit the bill by chance of their design.
...but then that specificity is something I'm finding hard to articulate... hence the thread! :p
Perhaps it's just a more full-on approach to elgonating a game's life-span, to use the base game as a starting point rather than a complete package that gets additional content over time. This is the kind of thing that defines GaaS to me.
That's why I prefer to use KPIs for specificity. Under a broad definition, anything with post-release content would fall under the GaaS umbrella, which makes the term kind of meaningless.
Specifically, GaaS from a KPI standpoint would require that the
service portion of the product (the post-release content) is self-sustaining, meaning it'll pay for operations. This includes both development and the cost of user acquisition, the latter of which is a major component to running a game
as a service as opposed to the old way of doing things, which blew the marketing budget at launch and didn't consider the problem of how to continually seek new users throughout the game's lifetime.
This is why I don't consider the Borderlands series GaaS in the true modern sense. They had bombastic launches, but never made an earnest effort to grow their userbase after launch. A true, modern GaaS built from the ground up launches with the expectation that after the initial post-launch drop in interest, engagement will actually steadily
increase. Older games, even if they technically fit GaaS definitions, do not have this expectation, and never run their operations with it in mind.
In more direct terms, a true modern GaaS uses a simple metric for success: cost of user acquisition is less than the lifetime value of the customer (CPA < LTV).