I just can't think of any reason why, if I were building any sort of business (tech or otherwise) I would incorporate google services (apart form google search ads, perhaps) into any aspect of it that I would deem "essential" or even "important". Even Drive & the associated sheets/docs/etc I look at with an eye towards "be ready to mass download the lot of it on short notice"
As an example, about two years ago they were building something that looked poised to compete with MS Power BI and Tableau. Not mature enough at that point, but promising. I poked around and saw potential for what it could offer if I pushed it forward in my work place, but I quickly realized it was a very risky career prospect: It might not work quite so well and sure, that's fine, we'd be trying it out on that possible basis. But it might work just fine, and we migrate some clunky "enterprise" reporting over, decommission some old iron, and then... ::Poof::! Google's system is gone, end-of-lifed with a 90 day notice and then I'm left holding the bag because Google is not a reliable partner for anything outside a narrow core of the selections.
There's a big difference between a paid (Google) service and a free one.
Google doesn't generally shutdown paid B2B services, but it certainly does have a record of instability on free consumer services (like this wifi thing).
Well, they may not exactly sjut it down in th same abrupt way if it's paid, but if you look at their recent changes to the costs if using their maps API, they're certainly willing to rewrite the terms of use for paid tiers that result in apps that were paying nominal fees to suddenly being charged onerous, crippling costs order of magnitude more expensive that require shutting down the service built on Google's tech.
So, no, paid services are not free from the fear of Google's drastic and suddenly project cuts or (as near as cuts) massiveness overhauls.
Google does turn down lots of (smaller) free consumer services, but the meme that their behavior around consumer products should translate into not trusting them for B2B is not backed up in their behavior or historical data around support. They behave similarly to all the other major companies.
For example the App Engine Datastore has existed for 12 years and still works, despite the availability of Firebase RTDB, FireStore, Cloud Spanner, Cloud BigTable, etc.
The one example is not a pattern, but I'm not familiar with all such moves by Google to know if there are others, but I freely admit there may not be. It is, however, an existence proof to show the possibility, which is enough to engender skepticism when choosing them as a vendor.
In my experience Apple is more considerate in such matters. They tend to API end of life with more lead time, barring some issue that has to be changed more abruptly either due to abuse, security, or bugs, which is not unreasonable. But even they don't approach the level of support guaranteed for traditional Enterpris software though, and Microsoft mostly fits that category. Sure they end support, but it's telegraphed years in advance. In large scale enterprise systems like ERP, a 10 year roadmap of product support and 3 year roadmap of features is pretty much standard. Albeit the products themselves are hardly best of breed, but there is a lot to be said for guaranteed stability of infrastructure, capabilities, and support, especially when you can bolt on your own custom needs (of course at additional expenses though.) Enterprise has these benefits, but you pay an extraordinarily high premium for that.
Also, special agreements with them (provided you're willing to pay even more) are often available to continue support even beyond the EOL of those Enterprise offerings. (My workplace had to do so for a major system for 3 years (it was a 40 year old system) due to delays in upgrades to the newer offering.
Sure it's unfair to expect Google to keep money losing products, but I think if they ate the cost to keep these sorts of things going a little longer for more graceful exits, they'd engender more consumer confidence and gain more in the long term.
It might have been that... They had a few similar forays into that sort of thing and at least one change of branding for the same offering when they gave it an overhaul, so I'm not sure. It may even still be around, but I just didn't have confidence in it's long term viability to push it organizationally over Tableau and ::shudder:: Cognos. (Cognos was mandatory, it was built on top of the the semantic layer of data from an ERP. A year later we realized we needed Tableau for truly modern data presentation, and used the semantic layer underlying Cognos as the primary data source for it, so it actually turned into a decent compromise. Now my push is for true data mining pipeline/ml development and deplo capabilities. Something like Alteryx, but there's a few solar industry-specific tools I'm looking at instead.
As an example, about two years ago they were building something that looked poised to compete with MS Power BI and Tableau. Not mature enough at that point, but promising. I poked around and saw potential for what it could offer if I pushed it forward in my work place, but I quickly realized it was a very risky career prospect: It might not work quite so well and sure, that's fine, we'd be trying it out on that possible basis. But it might work just fine, and we migrate some clunky "enterprise" reporting over, decommission some old iron, and then... ::Poof::! Google's system is gone, end-of-lifed with a 90 day notice and then I'm left holding the bag because Google is not a reliable partner for anything outside a narrow core of the selections.