User testing is often seen by business owners as an extra luxury to craft the perfect website, and unfortunately not seen as necessary to create a good website. It can be expensive, and add a lot of time to a project. User testing, though, is an invaluable tool which is often vital to the success of the web site or application. Without it, you never really know if your website was designed properly - you never know if it accomplishes what it needs to.

There's no reason basic testing can't be done for cheap. With Amazon Mechanical Turk, any website design can at least go through some form or user testing for less than $10! Spending as much $50 or $100 will enable you to go through a few iterations, and get results from dozens of testers.

Amazon Mechanical Turk is a service provided by Amazon which allows you to submit surveys or tasks to people, called a HIT (Human Intelligence Task). The point is to gain feedback or accomplish things which are not difficult for people, but too much to ask of a computer. So, simple questions, short questionnaires, easy tasks. Users in turn receive a small sum of money - sometimes as little as $0.05, if it's for a single question. You choose how much you're willing to compensate people in the hopes it's enough that people will sign up for your HIT.

For user testing or conducting quality assurance for a website, you need to be very precise about your questions, and break up the questioning into small chunks. Since you're not looking over someone's shoulder and users can be anywhere, you'll need to set up a functional prototype, or mockup in a password-free area that's easy to access. Pretend like you're talking to a computer - be very clear, and leave no room for human error or misinterpretation.

We had a client recently who had problems with a flash video player for their customers and needed a resolution, quickly. Is was difficult for us to isolate the problem since the client wasn't sure what questions to ask their customer. We were able to set up 2 alternate testing areas with the flash video player, and submitted 8-question surveys to 50 people. Within 2 hours, we had definitively isolated the problem area and realized what was causing it. We made the appropriate changes, and sent out the questions again to another 50 people.

In a single afternoon, we spent less than $100, tested over 100 people and fixed the problem area!

Check out this video we made on the topic: Clicky.