After 20+ product launches in ~10 yrs of doing this (apps, games, sites, big features), there's a pattern that I see repeat itself in the winning ones.
1) Team has to believe in the mission of the product so deeply that they fall in love with the problem, not the solution.
2) Execution is a function of 1) how well your team understands the problem, 2) how much your team cares about the problem, and 3) how quickly they can iterate. Don't add people to the team (especially early) that don't have a clear relationship to the problem. Despite skillset.
3) Better to find someone who cares deeply about a problem who needs help building it than it is to hire an expert at building but doesn't deeply care about the problem. Why? Core assumptions need to be challenged at every step. Learning is a great way to challenge assumptions.
4) Early team is usually 1-2 people that care about the problem so deeply they'll overcome any reasonable obstacle in their way. If you introduce process in problem-solving too early, you're severely limiting the creative process required to overcome obstacles.
5) Not 1 successful product I've ever launched was polished at first. Being scrappy helped me see that people cared so much about something they wanted to help improve it. Whether it was bugs in games, feature requests on websites, or emails for apps. Build scrappy with community
6) Big wins start with small, niche wins. E.g. in Runescape servers to be the first to release a scrappy version of a new feature the main game added. Only then expand on what is possible within that mission by listening to users. 2 years later we were the largest.
7) Trust is the biggest catalyst for speed and community. Not only do you create a good environment to work in, but your users will sense your team's bond and want to be a part of it. By investing in relationships and trust you invest in everything else.
8) It's okay to set high expectations together with people who love challenging assumptions. Doing this beforehand helps clear any misalignment down the road. But if you do that, you also have to be okay getting completely out of the way after the expectations have been set.
9) SEO and organic listing websites (forums, reddit, etc) are still the best way to launch things that sustain. Forget fomo. Medal started on a Rocket League subreddit. My 1st server started on a listing website called "Moparscape". Build what people care about. Show them naturally.