I wonder if they ever compared the price of AWS hosting and just renting a rack at a colocation facility.
One way of thinking about this is that "free" isn't "enterprise" and while you would love to have them convert, hosting free users on a rack with a 500MB Cogent IP transport contract can be had for < $2K/month. So $24K a year. Is it a "teir-1" data center? no. Does it have failure redundancy? no. Is there a risk of data loss in the event of a power failure? yes. But all of those things are what drive the costs up and are what make the "paid" plan, worth more than the "free" plan. Would it be good enough for people? Absolutely.
I'm pretty skeptical of this approach. There are generally two reasons to offer free tiers:
1- hope users will upgrade to paid plans
2- hope free users will spread name recognition through word of mouth to people who will pay
Generally speaking, "free tier" users are the worst- they demand more from support and expect to give nothing in return. Very few transition to paid users.
What happens if you give your free tier users a degraded experience compared to paid users? They'll complain that your service is slow, loses their data and is a terrible product. You're shooting yourself in the foot, because you'll get no fewer upgrades and a terrible reputation.
Even if you try to communicate that the free tier doesn't have the stability or performance of a paid plan, if I experience the free tier and am unhappy with it, you've raised the bar considerably for meeting my expectations for the paid plan. I'm more likely to pay for a competitor than pay for your product in the hopes that it'll improve.
No, I don't think that's right. I can name other scenarios:
1) I am at a megacorp. I want to rapidly prototype without going through purchasing / finance / bureaucracy.
2) I have a personal and a work account. I learn something on weekends on the free tier and use it at work.
Both of those have happened (with me, as the user), and led to millions of dollar of business to cloud vendors. In the case of #2, the cloud vendor has no way of knowing that the conversion came from a free tier. In the case of #1, I'm not even sure they tracked it.
I am astronomically less likely to use tools without a free tier, at the very least at the level of validating that the product is useful, and ideally, at enough of a level to get started.
The AWS free tier is almost exactly the right level, but would do better without expiration. It's not costly or usable to build anything significant in the free tier, but more than enough to get started. The one-year limit is an issue, since when I switched jobs, it was no longer usable.
Other scenarios I've seen:
- Early-stage startups. It's dirt cheap to offer free tier to someone with O(0) users, and most stay there. Those that hit the exponential growth curve more than make up for it, and won't have capacity to migrate. AWS courts these hard.
- Education. What people learn in school, they take to their jobs. Especially engineering software companies make hard plays here.
All good points, but I'll add that "free trial" is a very good hook, "free forever" is not.
So offering free trial periods (like AWS do for a year) is very effective, and very measurable.
Free Forever is clearly not self-sustaining, do one of us will inevitably be disappointed eventually, and i suspect that'll start by being you and end up being me.
I wouldn't make categorical statements like that. Free forever (without the promise) has worked well for many businesses, for github/gitlab, to Google, to quite a few others. A few models which have worked:
1) Limited to a single user (no team use);
2) Not have appropriate compliance for corporate use;
3) Limited enough to be free to provide and not viable for any real-world use (e.g. 15 minutes / 100 API calls per month);
4) Incompatible licensing (AGPL/proprietary dual licensing is an example);
5) Making everything on the free model world-public (MANY rapid prototyping tools do this)
The goal is usually to be adequate for nights/weekends personal projects, open-source projects, and internal prototyping, but to require customers who can pay to need to pay. How you do that depends on your business.
The higher-level point is that working from categorical points is not good business strategy. As with any tool, something like free, free forever, etc. can be good or bad, depending on how it's used, and requires a careful case-by-case analysis. If you remove tools from your toolchain based on categorical guidelines, you will be at a disadvantage.
The nice piece about the "forever" is that people change jobs every 3 years. I exhausted my trial tiers at several vendors in my first job (several of which switched to paying). Guess who's not being experimented with in my current job? All the vendors whose free tiers I exhausted, but whom I didn't adopt.
> I have a personal and a work account. I learn something on weekends on the free tier and use it at work.
But have you ever used some free tier, found it to be flaky or buggy and then still continued to upgrade to the enterprise tier because they promise you the paid tier is less flaky than the free tier? I wouldn't believe them.
#2 basically counts as a word of mouth reputation. #1 is essentially an upsell. Your other examples are the same, just the other way around.
In any case, if the free tier sucked, all of them would be unlikely to convert. People will remember that your product sucks, not that they had to pay to get decent performance.
It reminds me of the game "Game Dev Tycoon", where they released a version on torrent sites where after about a year of ingame time your company would go bankrupt because of loss of profits due to piracy. The retail version of the game didn't even include piracy as a mechanic, it was just added as a joke in the torrent version.
Still, Twitter and Reddit discussions about the game were filled with "don't buy that game. It's impossible to play for more than an hour without going bankrupt". Of course none of those users admitted that they pirated the game, and word-of-mouth reviews of the game became "don't buy it, it's terribly balanced / unfair"
Ah but the rub is that it would not be unreliable.
One of the things I discovered was that "reliable" in the eyes of non-engineers isn't the same thing. I have a NAS system for my important data because I know it is reliable, I know dozens of people who "have a copy on a flash stick" because they consider that reliable.
You can build "good enough" in a rack that will serve a LOT of clients. One layer of ZFS (1U + 3U) for storage, Five 1U servers, one 2U 10G switch.
In my experience that is 1 failure (that results in an outage) every 2 years at most. When I was at Google I (like everyone) had access to failure information about equipment in the "fleet." Motherboards do fail, but VERY rarely, disks are the worst offender but with double parity ZFS you have a week at a minimum to replace a failed drive. ALL the drive failures at Blekko announced themselves before they failed with SMART data.
So in addition to the colo cost you've got an engineer spending part of their time watching the rack, fractional head count cost.
I have a tiny VPS that costs less than $2 per month. I run a chat server on it. In seven years it has had no failures and one hour of planned downtime. That's better than any of the major cloud services. Simple stuff is surprisingly reliable.
I’m guessing that if you don’t join big federated chatrooms on your Matrix server, it runs absolutely fine on minimal resources.
One of the smarter things the bluesky team looks to be doing is to split the world into heavy-lifting versus personal servers, so folks don’t end up with their personal servers suddenly sucking up lots of resources just because a user views #wtf or whatever. We should look into this sort of
tiering too for Matrix, to eliminate the “Matrix is resource heavy” snark.
I don't federate. If I had more control over how much data other servers could dump on my server, other than restricting rooms with X number of users, I would. Synapse and Element together are heavier than any other chat servers/clients I've ran, with only one exception.
It is not a big problem, because my number of active users is fairly small (under 50), but I could see myself having to use a different server software, or possibly a different protocol all together, if it were to open up and grow.
And if that vpn had been hosted in OVH's Strasbourg datacenter, in 2021 you would have had a huge downtime and possibly total loss of data without a backup in another datacenter.
It is like driving under influence. Many people get away with it most if not all their life. Other die in stupid car crash alongside their pregnant wife mere days after their mariage.
One way of thinking about this is that "free" isn't "enterprise" and while you would love to have them convert, hosting free users on a rack with a 500MB Cogent IP transport contract can be had for < $2K/month. So $24K a year. Is it a "teir-1" data center? no. Does it have failure redundancy? no. Is there a risk of data loss in the event of a power failure? yes. But all of those things are what drive the costs up and are what make the "paid" plan, worth more than the "free" plan. Would it be good enough for people? Absolutely.