100% - I'm less and less inclined to build free tiers into my products these days, it's just not sustainable.
Microsoft figured this out relatively early on during their monopoly lawsuit which they cleverly settled by donating a bunch of Microsoft products to schools so that the next generation of the workforce would grow up on Microsoft products and become commercial Microsoft users in the workforce (and I've seen Apple use this same approach since then). What was ostensibly seen as a punishment by the legal system simply entrenched their position further.
And in gaming, there's an entire industry around "free to play" games that make billions of dollars, although I sincerely hope B2B tech doesn't take marketing inspiration from them.
Google as well recently - unlimited free G Suite (now Workspace) for schools was a massive draw a few years ago, although they are starting to turn around and monetise now they have a captive audience
There can be a lot of abuse, though.
In return schools teach their students Microsoft products in school and you can't really beat the pricing. Because running your own infrastructure will be more expensive especially if your school isn't big enough to have it's own IT department. Prices are rising though because you can't license box versions anymore.
So you now have the choice of paying for Teams, SharePoint online and not using them and licensing and running your local file servers, AD etc. or switching more of your infrastructure to the cloud and increasing the lock-in.
This is really worrying to me, but from a cost perspective you can't really justify continue running a lot of infrastructure, especially if your local government doesn't have a lot of funds in the first place. I'd rather spend that money on hardware that the students and teachers can use than pay for servers and licenses.
Of course you can:
https://www.microsoft.com/en-us/microsoft-365/p/office-profe...
https://www.microsoft.com/en-us/microsoft-365/p/office-home-...
To change this there would have to be political will to reduce the lock-in and maybe offer a state run cloud service.
If you decide to make people pay up, a competitor will step up and offer a free version, subsidised from VC or lucrative other businesses. And if the competitor isn't free, at least its cheaper: triggering a race-to-the-bottom, ending in "free".
In any market, for any service with potential, there will be a free, or at least cheaper option. Untill mono-/oligo-polies are established and prices go up, at which point the customers are "extorted" or close to it: all the losses up to now, must be paid from revenues now.
This is seen everywhere (from food delivery, via webhosting, to PAAS to SAAS) and a clear and present proof of why "free markets" aren't automatically efficient and need authorities poking around in them - or else they hardly work at all, let alone efficient.
Is there an opportunity cost in providing a gratis/marketing tier versus focusing more on paying customers?
>In any market, for any service with potential, there will be a free, or at least cheaper option.
Which is not bad at all. IMHO the only problem here are business decisions, slowly locking you into something. FOSS nerds prefer protocols over platforms, which is why Posix and Linux living up to it are so incredibly important. You are right about the markets being misaligned, which is why we have to be very sceptical about big corps buying into various FOSS-foundations.
Though I expect the same applies there too, only that "race to the bottom" is less of a negative thing. And could probably be "race to open-source". Where slowly all paid-for software gets FLOSS alternatives that -at some point- offer competing support/features/experience and therefore make the paid/proprietary alternatives more or less obsolete. E.g. while Oracle is still running and offering their database, on the whole, the world runs free databases. While Microsoft still sells billions of OSes, on the whole -including Android- the world runs Linux mostly. And except for OSX, hardly any other paid-for unixes survived. And so forth.
I do still consider offering a "free" tier, but with the limitation that there is no free support, and each support incident will cost them $200. But then I remember what I said above.
A $10,000 purchase for a small company with $100K revenue is going to be a huge expense. For a company with $100MM annual revenue it's a rounding error.
Free tiers are good if correctly sized compared to paid tiers: they help users get to try/know your service, they allow students and kids aroung the world to play with things, they help small companies grow; if your free/paid tiers are well done, those users who start with the free tier will grow to the paid tiers. If not, it's should be ok too, not everyone is going to need all your features.
Surely you can appreciate that "time" is not the only commodity that can be meaningfully (and reasonably) limited?
Some people will think that this is unfair, but I think it's great