Field studies are great for seeing real user behavior and pain points. It's also important to get out and test your concepts "in the wild" before you get too invested in code. That's what intercept studies are for.
You've got a neat prototype or early working code, and you want to check that you are on the right track. So long as your product won't blow up in someone's face, it's mature enough to use in an intercept study. This type of research work gets you many quick pieces of feedback while it's still early enough to make changes to the product cheaply.
Because you probably don't have much functionality, it's crazy to bring people in to your offices just for a five minute session. Instead, go to where the task occurs. This could be a waiting area, a coffee shop, a shopping mall, a colleague's desk, or on a train during a typical morning commute.
Going to where the action occurs has the added benefit of letting you see how your product would behave in its real environment. You'll quickly find out whether the assumptions you made are correct, and whether you accounted for people's preferences. This is especially important to do for mobile and tablet apps
The method itself is really simple. You just find people who look like they would fit your user profile, ask them if they'd be prepared to help you out, give them a quick task to perform and when they're done, you give them a small gift. Good gifts are things like coffee shop gift cards, or other easy-to-spend items.
As with all field or site visit type studies, you want to use active observation. Tell your participant what their end goal is, hand them the computer, device, or paper prototype, then shut up and watch. Save your questions until afterwards.
I wouldn't bother with video recording. Unless there's someone back in the office who really needs convincing, the video is just an additional hassle and it can act as a barrier between you and your participants. Easier to bring that person out with you than take video back to them.
- Make a sign “Get $5 for 5 minutes” or “Let me buy your coffee”
- Have a facilitator who interacts with the participant and an interceptor who is good at distracting your participant's kids or holding their shopping while they help you out.
- Keep it simple. 3 tasks maximum, 5 minutes maximum.
- Find a group of people? Do co-discovery, where one of them drives and the others look over the driver's shoulder and say what they should do next. This way you get to hear what they'd all do and why.
- Be prepared to be ignored by the majority of people. They probably think you're selling religion or insurance.
- Take and wear company ID so you look more trustable.
- Clear your work with building management at your destination - shopping malls are private property, and coffee shops might think you're scaring their customers away if they don't know what you're up to.