| You're worrying about a business with a estimated minimum profit margin of 79%, using the most pessimistic assumptions I could think of. I wouldn't be worrying yet. In 2017^: revenue was $259k, costs were estimated $17k, archives 31.8TBy, 126M URLs. In 2020: revenue was $212k, costs were not specified, archives 82TBy, 192M URLs. Archives are 2.58x higher, URLs are 1.52x higher. So taking the higher of the two multiples and pessimistically multiplying the 2017 costs of $17k (the cloud has gotten cheaper but let's ignore that), we get $44k in costs (this is probably too high by 2x), leaving a profit of $168k. Users dropped from 29K to 19K, so you should start worrying when users reaches 5K, assuming that costs scale linearly with data (which they definitely do not). In all likelihood, you probably don't need to worry until users reaches 2K. At current rate, that's probably about a decade, assuming ten more years of no changes to the site (such as mobile support) and etc. ^ https://blog.pinboard.in/2016/07/pinboard_turns_seven/ |
A couple of data points to help you calculate:
* active users is not the same as paying users, it's people who did something in the last month
* costs in 2020 will be something like $17,000 for hosting plus $8,000 for hardware and a few thousand bucks in travel/hotel costs to the colo. So say $30K.
* the reason hardware costs are high in 2020 is... massive expansion, so maybe people in this thread are missing something