- Bottom line, the game is not particularly fun, novel, or difficult. You'll notice that games like Flappy Bird, Agar.io, Papers Please, etc. are at least one of the three. Bist.io is not the type of game that will ever go viral.
- You spent money on AdWords but didn't spend money on sponsored Youtube, Instagram, or Twitch content? To me, this just shows a fundamental disconnect with your target audience (gamers).
- You should never polish a multi-player game until extensive play-testing. Single player games are a bit different, but with multiplayer games it's the mechanics that sell it. If the mechanics aren't fun, polish doesn't matter. Case in point: H1Z1 and PUBG. Awfully optimized, and generally terrible lag, but still massively fun and massively successful.
- Too much premature optimization. Having your servers blow up is a good thing. Don't worry about preventing the blow up until you get there. In other words, you spent too much time on load balancing, etc. This is the canonical reason good engineers make bad business decisions.
Going into something (even games) to make the big bucks is perfectly fine, imo, but, in short, your game just isn't very good and you have no idea who your audience is. You added "social" buttons because it's the "thing you're supposed to do" but omitted Twitch and Steam? You're not on IndieDB or TIGSource?