A lot of great startups did not do user testing while making product. They give more thought to having the basic focus right- the one thing that the product does- it should do that well. BUT after they launched it, they started taking user feedback and organized informal testing sessions. Here are a few case studies for the same.
Blog of meebo (relevant to “posting it to a forum” reply by Tara)
“Seeing something in person is always different from hearing about it. When users write in with feedback, it’s really useful, but sometimes we wish we could be there in person to see a bug or witness the smaller inconveniences that people tend to gloss over.
So last Thursday, we hosted our first Usability study and invited nine volunteers to use meebo while we took notes and gazed anxiously over their shoulders.We had a great mix of volunteers: a 14-year old, an avid MSN user from Holland, and even a husband and wife duo. We asked them to perform tasks like removing buddies, joining a group chat, adjusting sound preferences, and just demonstrating their typical meebo routine. We’re still compiling notes (pages and pages of them!) and consolidating findings but there were a few things that immediately stood out.”
“Gmail was used internally for nearly 2years prior to launch to the public. They discovered there was approximately 6 types of email users, and Gmail has been designed to accommodate these 6.”
They use the 20% / 5% rules. If at least 20% of people use a feature, then it will be included. At least 5% of people need to use a particular search preference before it will make it into the ‘Advanced Preferences’.
So, for me this means that there are multiple ways of improving a product while taking user’s feedback. Expensive user testing is not required. Thats what I am doing for this product. 5 users for each profile are great for initial feeback.. lets see how that goes.