Some people engage in risky behavior for fun – this makes no sense but luckily, it's mostly unplanned and pointless. The real topic of interest is people taking risks deliberately.
We can imagine the world as being a sort of Market+ – there's a lot of goals available in the world, and many aren't even related to traded goods and services, but the same insight holds for almost all deliberate behavior: the more crowded the field is, the harder it is to be world-class, and the less valuable your efforts are taken to be. If there are ten million people trying to be the next best American poet, there's a sort of strange sadness to following that pursuit. If there are only ten, but the value you get from poetry is similar, you're on a path to the stars.
In this model, you'd think about everything coupled with risk. Should you start a remote piecemeal consulting accounting firm for companies considering ICOs or a restaurant? Idea quality being equal, even risk being equal, you should prefer the accounting firm – it is a bit weird, and perhaps not quite like other businesses, but because it represents an illegible risk, it is less tempting. At the margin, you'd expect it to represent a much better risk profile – even if it's unknown how risky it is – than the restaurant. Everyone knows those are risky, about how risky restaurants can be, and the factors you can control. Strange new business models have unknown risk, that can't be as easily managed – but that doesn't actually entail more work, often, just a higher willingness to have done something that has failed.
Restaurants also provide lots of excuses – location, location, location, immature Yelp reviewers, strange market forces like youngsters insisting on delivery but your food being best very hot and fresh, local economic concerns, macroeconomic concerns. If it is a franchise restaurant, it provides even more excuses for failure: hey, look, it wasn't even my idea! If you don't run any risk of embarrassment, people who dread being seen as a failure will compete with you, but if you risk being humiliated by failure, you'll always have fewer competitors and probably much better margins.
In many ways, all other things being equal, you should prefer an idea that has only and exactly the things you want out of the venture. Sometimes that will be money – there's a lot to recommend having money, and it's possible fear of being seen as greedy limits people's honesty about that, and thereby limits their happiness. Sometimes it will be some composite social good, or connections to a specific community.
But once you've tallied all the things you care about, it might be helpful to brainstorm, before you take your risk, ideas that specifically don't have benefits you're willing to forego. Particularly common benefits – doing boring work, for instance, can be a path to success. I mention things like shielding from embarrassment because it is by far one of the strongest emotions we feel, and is a profound motivator. I suspect the biggest margins are in accepting more shame.
What is the least cool tech startup? Tech's a growing industry seemingly obsessed with startups looking cool. I'd anticipate crazy growth and crazy margins in the most tragically uncool corners. What is the tech company with the least growth potential? If you just want a normal sized business, this sector also suffers from obsession with rapid growth. Appropriate for some, but lots of money pushes the most dangerous competitors into hyper-competition around the scale of growth. Local bakery-sized businesses would be a stroll in the park, in comparison.