Guerrilla research is a method of user research that is perfect for teams that want to move fast through a digital design project, test ideas early and often, and gather meaningful user experience insights without the overhead of more traditional methods of research.
It doesn’t replace traditional or deeper forms of user research, but it does let a team quickly understand what does and doesn’t work in their products - whether software, mobile apps, or web apps. The best part? Guerrilla research can be performed anywhere you would like with little preparation, in a tiny amount of time.
Guerrilla research is perfect if you’re looking to validate or invalidate hypotheses around functionality that don’t require domain knowledge. In the past, I have used it to test search and filtering functionality, multi-select, and sign-up flows – none of which required them to have deep knowledge of a specific topic.
What You’ll Need
During these research sessions you will be out of the office in a public place, so you’ll need to travel light and plan ahead for what you will need. Here’s a handy list to get started, add or remove as necessary.
Not including running a pilot test with colleagues before venturing out, you should set aside one or two hours to perform the research.
- A person who can lead the research sessions alone. This person should feel confident guiding a research session on the fly, coming up with questions at the drop of the hat to get an understanding of someone’s experience using a design.
- A prototype of any fidelity that has been tested with someone else (e.g., a colleague) before taking it out of the office.
- A brief for the research session so you know what you are testing, the flow you’re taking through a prototype, and something to keep your session focused.
- A laptop or any other equipment you might need to test the prototype. This should be equipped with software to record the interview (I tend to use QuickTime).
- A place where you feel comfortable approaching strangers and asking for a few moments of their time. Student unions, coffee shops, or even launderettes are great for this as people tend to be relaxed and in a positive mode.
- Bonus: Have some incentives prepared to gift people once the research has been completed. If you speak to five people, give each of them a $5 gift card or set up a tab at the coffee shop.
Session Format
Each session should last between five and 15 minutes. This is a perfect amount of time to not bother the participant, but enough time to get the insights you need to keep moving on a project.
Here is an outline of an ideal session:
- Walk up to somebody and ask if they have a little time to help you with something, make sure they know you aren’t selling anything, and provide credentials (e.g. a business card) to make them feel comfortable.
- Don’t be pushy, and only proceed if the person accepts. If they decline, then thank them for their time and apologize for the interruption.
- If the person has the time, show them your laptop and set the stage. Explain that you are looking for feedback on some designs, and that they aren’t being tested for their ability to use a computer. If you aren’t looking for feedback on design and colors, make sure you tell them so.
- Get the permission to record the prototype and audio, letting them know that there will be no other kind of recording (such as their face) unless that is a requirement of the research.
- Do not write any notes during the interview, this is why we are recording the session.
- Begin walking through the research session, asking them questions required to get insights you need in the research session. Refer to the brief or discussion guide if necessary.
- Upon completion, make sure you give them a sincere thank you and any incentives you might have.
If the feedback you’re receiving during the session feels unhelpful, try to steer the session back toward the discussion guide, otherwise thank them for their time and move to another person when it feels appropriate.
When Sessions Are Complete
When you are satisfied that you have enough information from the search (maximum of six participants), take notes of key takeaways and interpretations by using the recordings of the feedback.
Note anything that feels important from the research, making sure it doesn’t include any solution for the problem at hand. Research is about validating and invalidating you work, not coming up with solutions on the fly. For example:
Four out of five people could not work out how to select more than one thumbnail.
Finally, If you’ve got wins there make sure you celebrate them with your team.
Choose When to Use Guerrilla Research
User research is less useful if you’re using the wrong method at the wrong time. Be careful when choosing how you validate or invalidate your assumptions, so make sure you only use guerrilla research when it is an appropriate methodology.
Guerrilla research is an effective method for understanding how real people interact with your product. The correct methods and an analytical mind will help you gather useful insights on what is working well and what still needs work. When done right, guerrilla research is an essential practice for anybody regardless of budgets or timelines.
Further Reading on Guerrilla Research
- The Pros and Cons of Guerrilla Research by the IDF
- What is guerrilla testing and how do I use it by Elizabeth Chesters
- Guerrilla research: quick, not dirty by Foolproof
DockYard is a digital product agency offering custom software, mobile, and web application development consulting. We provide exceptional professional services in strategy, user experience, design, and full stack engineering using Ember.js and Elixir. With staff nationwide, we’ve got consultants in key markets across the United States, including San Francisco, San Diego, Phoenix, Dallas, Detroit, Miami, Pittsburgh, Baltimore, and New York.
