The way I'm coming to understand, there are multiple problems to address, some of which emerge at different stages of site / business development:
- Initial growth: finding and cultivating participants. Note that many of the more successful / notable services essentially started with an extant cohort, notably Usenet and Facebook, both of which originated at selective-admissions universities.
- To borrow a term from AI: alignment. You might also consider "purpose" or "mission statement". Essentially, what are you trying to do, who are you trying to serve, and how do you keep committed to that path? A huge problem with many past online services is that they started user-focused (or at least appeared to), and ended up investor-focused (if even that), often sacrificing or cannibalising early ideals. (Reddit, Facebook, Twitter, Slashdot, ...).
- Business model. How do you keep the lights and servers on, pay bills, meet payroll, etc., etc. If you've taken third-party funding, what's the Devil's deal there?
- What I call "hygiene factors": how will you address all the negative side-effects of growth, once those emerge? There's a bunch of trust-and-safety issues, problems of managing a growing business, investor/creditor relations, and all that jazz. One of the unique problems facing any media organisation is that once it reaches a level of prominence, it becomes an attractive target for manipulators and value-extractors, whether by the putative operators or third parties, and addressing that risk is a considerable challenge. At the same time, that's not the most significant problem facing the organisation at its inception.
Another intrinsic challenge of the start-up world is that starting up is easy, it's the not-shutting-down, and the staying true to mission, which are hard. And the list of community-based discussion communities which have succeeded in avoiding those pitfalls is exceedingly short, if not nil.
Transparency and accountability ... seem pretty empty to me these days. I've noted before that even such mechanisms as B-Corp status end up being meaningless if they can be easily subverted, and/or the enterprise fails regardless. Ello, which I'd participated in significantly and with some initial enthusiasm, failed to live up to its early promise. Imzy, another Reddit alternative ("kinder, gentler") also fell to a number of self-inflicted wounds (in part attempting to solve a later-stage problem in its early, inception stage; also acquiring a highly toxic initial member cohort).