Can you identify a business problem a taco truck has? (I'm assuming taco trucks are something within your experience. If not, assume I said "the last restaurant you ate at") Even if you know nothing in particular about taco trucks, you can go the the reliable old standby that absolutely every business in the world feels its revenues could be higher and its costs could be lower.
Let's focus on revenue. Do you understand how a taco truck generates revenue? They sell tacos (and some other stuff), at a particular price. If they sell more tacos, holding price equal, their revenue goes up. (Take on faith that tacos are incredibly high margin and that selling more of them is a win, OK? I strongly believe this to be true for the typical taco truck. If you don't know what high margin even means, that's fine, because it doesn't change this analysis.)
So our problem is now "Sell more tacos, with software." Can you do that? If not, ignore the software bit -- can you dream up ways to sell more tacos? Can you just brainstorm twenty of them?
+ Get more people to hear about our taco truck.
+ Always ask people "Would you like a second taco for the road?"
+ Sell tacos at a better location.
+ Sell tacos to companies in big batches, then anchor your retail business at the company's location for a day.
Now look very hard at your list. Is one or more of those amenable to being implemented, in whole or in part, in software? I suspect the answer would be yes.
You don't have to be bitten by a radioactive spider to make businesses money, and it equally doesn't require superpowers to do that with software. If you absolutely, positively feel you can't do it, I strongly suggest working in industry for a year or three. Any industry will do. Applying technology to discrete problems is a learnable skill.
This is making me think why stop at software solutions? Why not just solve business problems in general, but use Ruby on Rails as a tool to solve them? Maybe some problems won't even need Ruby on Rails!
Another questions comes to mind: how would I sell a service with such unpredictable outcomes? If I sell software, then I can almost guarantee certain features. But if I were to promise "more customers" or "more sales", I wouldn't even be 100% sure myself if my methods will work. How would I market a service with an unpredictable success rate?
I personally would have jumped on the most obvious problem, which is that your potential customers have no idea where to find you. See https://twitter.com/TheTacoTruck for a software solution to that. There are much better possible solutions from the customer's point of view.