I got a surprisingly high profit-per-investment-per-second in farmer's-market type goods at first, so I'd have to recommend that as your first store. Try to stagger your research per building, so you always have some 30min research, some 2hr research, some 8hr research lying around--that way, when you are away from computer, you always have something useful to time correctly.
Find a good, useful one-step manufacture that's useful to a lot of people, and dump research and expansion into it when possible, once you've got your store running. Like...I'm slowly moving to take over the cheap vanilla extract market (it'll take quite a while). You could probably make a fair amount on gelatin, or flour, or cocoa butter, or sugar, or something like that. Hell, if I wanted to start over right now, I would devote all my resources and research solely to artificial vanilla extract. You'll get a lot of people who want good tech from the second step of development. Why research wheat and flour when you can buy the flour? etc.
Look for cheap >Q16 water every once in a while...I just picked up like six million Q22 water at $0.17, which makes me happy. Cheap eggs, cheap milk, also good stuff (if you're me).
Personally, I'm not worried about niche /store/ markets. Aim for niche B2B markets. Find a tech that's used in a ton of stuff...like, nobody buys my orange extract, but tons of people buy my equivalent-tech vanilla extract.