This is very true! When we finished building Localist, our first partner was Johns Hopkins University. While the core of our product was pretty solid, the intricate changes related to workflows and UI came almost entirely from the school's feedback. We wouldn't be nearly as successful as we are today without their guidance.
Coincidentally, we're doing exactly this in my new company, since the last couple of months.
Would love to get the pitfalls of doing this and what we need to be careful about.
Many others have said this, but another way to make sure you don't fail at launch is to market before you write a single line of code. Start by building an email list. Plus, then you'll know where to find that first customer/co-founder as discussed in the post.
this is great advice, but I would add a caveat. Don't become a custom dev shop for that customer. Listen to your custoemr but make sure he remains a customer and not your product management or test/QA function.
The difficult part is product development. Just because one is a software engineer doesn't mean that one necessarily has good ideas for new products. It would be much like a great cook doesn't necessarily know how to create a restaurant. The skills needed for entrepreneurial success are different than those needed for software engineering. In reality, starting a great company is much closer to starting a restaurant than it to simply making quality meals. One has to know what to make as much as knowing how to make what you make into something people actually want to buy.