Here’s a question that I often encounter when helping companies conduct user research: Designers, product managers, and sometimes front-end developers will come to me and say: We’ll do some guerrilla/DIY user research to validate this design idea.
Just in case you mistake this for Chairman Mao’s Guerrilla Warfare — guerrilla user research refers to informal, small-scale, on-the-fly user research, typically conducted by non-UX-professionals to validate product design.
It’s a big part of the so-called lean UX process, an approach that has gained much popularity with the advent the lean start-up movement.
Things to Take Care of in Conducting Guerrilla User Research
I’m a big advocate of lean UX research — conducting user research in a responsive and flexible manner to support rapid product iteration. Therefore, my answer is “Yes” to guerrilla or DIY user research. On the other hand, there’re a few things you need to take care of before doing this:
1. Talk to the right users.
The biggest challenge for casual user researchers is to identify and find the right users before you start the interviews. It’s easy said than done so we need to be quite intentional in finding those people. For instance, if you’re developing a solution to make your eCommerce site more social media friendly, you need to talk to people who already use social media, and ideally you talk to both people who use social media with online shopping and those who don’t. Figuring out whom you should talk to heavily depends on your products and the questions you want to answer, so make sure to do a great deal of thinking on that.
Talking to the wrong users will give you misleading insight, and you’ll end up at a worse place than you started.
2. Don’t ask leading questions
I think it’s a truism that everyone knows, but it’s not easy not to ask leading questions if you’re the one that develops the product and hold strong opinions about it. It’d be great if you develop a moderator guide which can help you avoid accidental leading questions.
3. Create tasks that are meaningful and natural
This only applies if you are evaluating usability, which requires you to create tasks for users to perform. The tricky part is, it’s often the case that you will use tasks that are not natural to try to focus on a very peculiar element of the user workflow. For example, you might have users add an item to their shopping cart and check out. But this task is too artificial, as users are forced to add an item that they’re not interested. Therefore, the insight gathered will not shed light on how users naturally check out items. Instead of this, you can try a so-called “compelled shopping” task, in which the user takes time to find something he or she really want to buy and then check out. That’s how you get real user insight.
4. Look at end-to-end user process
Many folks that do guerrilla user research are responsible for a particular product feature, and therefore they only focus on that piece of user experience. But as we all know, user experience is heavily influenced by the context. Looking at a touchpoint in isolation will never give you a holistic picture. So you should look at the end-to-end user process, starting from how they discover your website — maybe through Google search — all the way through when they leave your site — you can ask what they would do next and why.
5. Gather users’ relevant background information
Different types of users behave differently. For example, tech savvy users might find a product easy to use while less savvy users might find it otherwise. Loyal eBay buyers perceive value differently than newbie eBay buyers. eBay sellers and eBay buyers see buyer protection differently.
Good user insight always come from context-rich understanding. Therefore asking some background information such as who they are, what websites they typically visit, how tech savvy they are, and prior experience with similar products goes a long distance in giving you the context in which to interpret user findings.
For example, when we helped with UX design on the 4evercard app and the Complaint app, we interviewed users by starting with understanding their background and therefore were able to get the correct context of the user input.
Misleading Insight Is Worse Than No Insight at All
In a word, it’s totally advisable to do guerrilla user research because it just works well with your lean product development process. But you need to make sure that you take care of the aforementioned points. Remember, misleading insight is worse than no insight at all!
To learn more about this topic:
Please read my article on how to conduct lean UX research, published on UX Magazine, for more information on the topic.
By Frank Guo