By my senior year in college I was working full-time on Hannon Hill both building and selling content management software. Just down the road there was a prestigious business school where I had networked my way into the IT director that was in charge of the website. After a great meeting with him, he said, “If you add kerberos for authentication, we’d be very interested.”
Naturally, my “happy ears” perked up and I thought that if we added that feature they’d be a customer. I emailed him after the meeting and set up a time to meet two weeks later. The night before our scheduled meeting to walk through the new kerberos module, I still hadn’t gotten it working (extra C++ code to go with the core PHP app). Undeterred, I stayed up the entire night writing code (literally!) and walked into the 9am meeting with a beautiful kerberos module. The potential buyer took one look, said they’d chosen another product, and wouldn’t be moving forward as a customer. I was devastated.
Positive feedback from a prospect doesn’t equal a sale (verbals are for gerbils). Ask for the sale before building the next set of features. Ensure the prospect is truly buying before committing more resources. Don’t make the common entrepreneur mistake: check for happy ears and get a sales contract before moving forward.
What else? What are some more thoughts on happy ears and sales contracts?
David Cummings on Startups 
It’s not a sale until the check clears.
Classic mud pie. When an entrepreneur leads with the product and not what the prospect is reaching for, leads to indifference.
People will be nice and say that they are interested, but that’s not real.
Just like a parent that tells their child the plate full of mud is a delicious pie.