Start making a list of what features it should have - there is no basic social network feature set, so that's on you to decide. Making a really big list and then ruthlessly crossing out anything that's not absolutely necessary for the app to be worth using can be a helpful method. (You can always add more in updates if the first version - the minimum viable product - succeeds.)
Keep in mind that both social networking and dating are affected more by network effects than anything else - people don't usually try brand new social networks with few users (many are created and fail without you ever hearing about them), they use Facebook even though they hate the UI because that's what everyone uses, and only seriously look at dating apps that have tons of users, or a moderate user base with some kind of filtering they find useful (like dating "for Christians", "for vegetarians", etc). Most of the current big social network sites got off the ground by going through rounds of venture capital investment, and spent insane amounts of money on marketing to users without making any profit. I don't know as much about dating, but I'd expect most dating sites that aren't super niche probably took the same path.
Paul Graham of Y Combinator has some good writing on this stuff. If you're still sold on doing this sort of business, I highly recommend reading his articles.
