In my opinion, creating a few test scenarios (tasks, if you will) and asking a group of people to do them is a great way to get some quick information. Small groups of people are a good start, but larger groups are more effective — allows you to filter out the “one person” issues from the “most people” issues.

I generally shy away from doing a test with less than five testers. The signal to noise ratio isn’t very good that way.

But I think that even at 5 to 10 people you’re able to walk away with useful information, as long as you take it with a grain of salt. You don’t need a large budget to get good information!

Those are all great points that you make. I think the difficulty sometimes is recognizing in some areas what you know that you don’t know about the people who might use your site.

What do you think of low budget usability testing? Do you think that there’s a value in getting 1-5 people who might be among your targeted audience, and watching over their shoulders as they try to perform some task on your web site?