Wikipedia's success can, in my opinion, be directly attributed to its noncommercial nature, which is enabled by its "many reads, few writes" model. Websites of

Wikipedia's success can, in my opinion, be directly attributed to its noncommercial nature, which is enabled by its "many reads, few writes" model. Websites of that sort are much cheaper to run than more interactive web applications, and are therefore better suitable for a non-profit model. For the vast majority of user requests, you just serve a static page from a cache. I suspect that most requests don't even touch Wikipedia's PHP code or database. The same is true for Hacker News (which entirely runs on one or two machines) and Stackoverflow, which runs on a dozen or so. Wikipedia is a lot bigger so the number of machines is probably somewhat bigger too, but it's still not comparable to social-media websites or even messaging apps like Signal.