A few years ago I worked as a designer at the corporate headquarters for a national real estate company. We had plenty of budget for design and marketing, but basically no budget for usability testing in spite of its importance. Here’s what we did.
Our engineering, design, and marketing offices were on the fourth and top floor of the building. On the first floor was a coffee shop with a few tables. I’d cultivated a good relationship with the owner and several baristas in my time there, and this proved to be an asset.
The research we needed to do was around people searching for homes to buy. It’s an activity that a lot of people have done, want to do, or can relate to whether they’ve bought a home before or not, so that helped.
I started with setting up a laptop to capture screen actions. I also had available a recorder app on my phone should it be necessary.
I talked to the owner of the coffee shop and explained what I wanted to do - buy people coffee in return for letting me watch them use our website, and competitor websites, to try and find a home. The baristas were happy to let me put a sign on the counter offering free coffee in return for using our websites, and at the end of the session I tipped generously.
When possible I took someone else from the team with me to record what happened. One time I took an engineer with me and encountered disaster when he began defending the product, despite my having explained to the engineer that we were there to observe only beforehand.
On the days I worked alone I used the recorder to capture answers to questions with the user’s permission.
You have to grade on a curve using this method as users are not pre-selected according to qualifications. That said, the site was designed to be easy and simple to use, and we got lots of great data.
I used this method for a year. We learned a lot. In case you’re wondering, users looking for homes to buy wanted to see big photos and see them fast. They also wanted to see results on a map showing the proximity to schools and other places they frequent.
There's an old and excellent book on this method called Don't Make Me Think by Steve Krug.