Do you want to market your product independently? If no, why bother? Just sell it and pocket the money.
Does your customer plans to sell it? If yes, and he is a small business (as it feels from your posting), you will be probably the best candiate to handle final sale and installation anyway, so keep your options open and do not make him angry.
After trying to use it, unless client is total moron, he understands that it's easier for author to maintain the product, than for somebody else, who needs to learn it. So your position is safe. And if he is a moron, you got some money from it, so deal is not all bad.

If he will propose to market it together, you can split the expenses, risk, and the profit. It's fair, believe me. If he is a person capable to go out and bring you a customer ready to sign a check, so you just need to explain details, install and train somebody, think about it from following point: You do *not* need to go out, try 100 times to sell, and get rejected 99 times before making sale at 100-th customer. You can save all this time, all this frustration, tinker instead with your program, and have fun coding.

If he wants to install it on many compurers, I will not bother, if you are expected to train and support only one installation. If he expect you to support many users, you need to discus that - and as a honest misunderstanding, but not attempt to cheat each other. Your contract for maintenance should mention how many users/installation you are expected to support.

To make importance of your work stand more clear, do not overdo documentation for first installation... ;)

pmas
To make errors is human. But to make million errors per second, you need a computer.