Unlike Slack, Mattermost is open source...
Teams suffer when SaaS companies lose focus. Since it’s open source,
Mattermost can endure through its community. If quality declines,
anyone can fork the code and take stewardship.
Open source and focus are the wrong arguments here.
(a) Mattermost is built by a video game company, so there is no focus to start with.
(b) Moving from a SaaS product to a self-hosted open source component runs against the process of commoditisation so is ultimately only ever going to be of niche value. After all, why spend scarce and expensive engineering time setting up, maintaining, and potentially having to fork a product, when you could just pay $however-much/month and not have to worry about it.
The problem of SaaS is having your company data stored by someone you have to trust... So I reckon it depends on your use case: if I'm a small company who doesn't want to bother running/maintaining the service then I'm more than happy to hand it over to someone else and trust them. If I don't want to trust anyone with my data I'd rather have something simple to deploy and run myself.
The same way that having a closed client prevents you from customising it to fit your needs but can be helpful if you had no intention to do so in the first place.
Yeah, there'll always be companies like that but as more and more large orgs, government departments, etc. migrate all their email over to Gmail, I'm starting to consider that as a niche. Those orgs are also disproportionately likely to want to build their own thing because they have a hope of making the economics of scale add up.
True for big corps, although it's quite strange given they're usually the most paranoid ones on security.
But I was more thinking about techie SMBs to whom running a chat/collaboration server is not scary, better in terms of flexibility and cheaper than paying someone to handle it.
(a) Mattermost is built by a video game company, so there is no focus to start with.
(b) Moving from a SaaS product to a self-hosted open source component runs against the process of commoditisation so is ultimately only ever going to be of niche value. After all, why spend scarce and expensive engineering time setting up, maintaining, and potentially having to fork a product, when you could just pay $however-much/month and not have to worry about it.