some thoughts on the end of cohost

Written on

so it’s over, more or less. some brief thoughts on things that went wrong:

  • the worst decision we made was to get into social media. originally, we were planning to ship a subscription management site, like patreon, but another group of people got into that space with a more radical take on it, and we didn’t want to be the Big Bad Corporation coming in and stomping on their feet. initially we talked with them a little bit about joining forces, but we couldn’t make the money/organizational side of that work for various reasons, and since we had more resources than them at the time, we decided to go build a social media site and then loop back to subscriptions later (which never happened.) the problem is twofold, really:

    • first, it’s a lot more obvious how to cover operating expenses for a subscription management platform than a social media platform.
    • second — and we really should’ve seen this coming — a social media platform is a lot more operationally complicated than a subscription management platform, and we ended up being perpetually swamped with work just to tread water.

    this was a deeply bad decision. we’re very happy that comradery is going to outlive us, though. go check them out. here is another link to comradery if you aren’t getting the hint.
  • the second worst decision we made was to get into social media when we did. we saw that social media was headed for a weird place a while before 2022, and our goal was not to make a “twitter replacement”, but certain billionaires’ incredible job entering the space reoriented everything around finding answers to the question “what will replace twitter?”, and dropped a bunch of people on our heads, with very strong expectations of how social media works and very little patience, all at once, right around when we were first publicly launching. we had some scaling problems already, and it made all of them worse and resulted in a lot of our addressable base getting a bad first impression of us.
  • covid did not help. we haven’t done a full accounting but we probably lost at least 10% of our total developer time to sick time and decreased productivity from long covid directly — on top of the fact that covid and post-covid inflation, coupled with our salaries being tied to cost of living, caused salary (the company’s largest expense) to increase by roughly 20% per employee.
  • knowledge is avoiding broad-ranging, opinionated frameworks because you know that using them will create a bunch of tech debt from them doing things the wrong way that you have to take back later. wisdom is using broad-ranging, opinionated frameworks because you know that not using them will create a bunch of tech debt from you doing things the wrong way that you have to take back later, and also take longer. the real choice is between getting the project built out and having the money to hire people to fix the tech debt, and not getting the project built out and not having the money to hire people to fix the tech debt.
  • fun cohost fact: the entire company was run for five years on outside funding amounting to a little less than the budget of this surveillance camera shaped like an owl. (no offense to the people who built ulo, this was just the first roughly comparable kickstarter I found.)

Tags

Leave a Reply

Your email address will not be published. Required fields are marked *