Guerrilla research is a technique borrowed from the marketing industry which takes usability testing out of the lab into the “real” world, where participants are recruited on the spot and asked to share their opinion for a few minutes. When conducted properly, it can be a valuable, low-cost and time-saving way to gather feedback on your product or concept; allowing researchers to communicate with and observe consumers in a public environment rather than an artificial setup.
We have recently conducted our first set of guerrilla testing at a public station in Berlin Alexanderplatz, asking people who had just arrived from a trip their feedback on a concept for a new app feature, while observing them interact with a very low-fi prototype. While we were able to collect very interesting insights, we have also encountered many challenges in conducting this type of research method that need to be addressed in the future. Here are our learnings:
Tips for conducting Guerrilla UX Research
1# Screening participants
When you do guerrilla you cannot really be too picky with your participant selection. There is simply no way to setup a “screener” survey or start interviewing people and then change your mind when you find they are not a good fit.
‘5€ for 5 minutes of your time, but only if you have Android and use our app!’ ‘Hey, mind answering a few questions, except if you are a student or unemployed?’
I believe this is two-folded. First of all, better set your requirements to a minimum. Guerrilla testing simply is not suited for testing with users who are all e.g., from Portugal, aged 20–35, English-speaking, studying at university and owning the latest iPhone. What guerrilla testing is especially good at is focusing on one (or two) or these requirements and finding as many people as possible with that one thing in common. Which brings me to my second point — setting up your ‘screener’, e. g. the where and the when of the test. Are you looking for Portuguese customers? Go to Portugal (or, well.. Portuguese restaurants, meetups, events) to increase your chance of finding them. Students? Go to a school. Working people? Catch them on their smoke break. Travellers? Bus stations, train stations, airports. iPhone connoisseurs? Apple stores there you go!
#2 Approaching Participants
this one is always a challenge, even more so when people have no idea who you are or what you are doing and you only have so much time to make a good first impression. During our test at the station, we found that many people were in a hurry and they dismissed us within 30 seconds from approaching them.
we had much higher success approaching people who were just ‘hanging around’ by themselves, smoking a cigarette or reading a book, or waiting for a friend to pick them up at the station. It was important to us to make sure they spoke English and were willing to spare a few minutes to answer a questions before mentioning our incentives (i.e. travel vouchers and a mug). We found that many people were willing to help even without an incentive at first, but they start getting worried once you mention things like ‘prototype’ and ask for their voice to be recorded. Plus, everyone loves to get free mugs — so make sure to have something to give in return for their time.
you technically don’t need to collect signatures from people for just talking to them for 5 minutes. You may be tempted to record their voice, though, and in that case it’s important you ask for their permission in advance. At the same time, you don’t want them to have to go through having to read and sign papers, which could take more time than the interview itself.
we wanted to show our participants a prototype and record their interaction and what they were saying using Lookback on one of our phones. On another phone, we setup a typeform containing a shorter version of our standard consent form — the participant just had to type in their full name and select ‘I agree’.
#4 Maintaining your cool
following best interview practices — not so easy when you are outside of the lab and both you and the participants are likely in a rush. We ended up talking too much and asking many leading questions. But we were also be able to adapt our interview style to each participant (which we needed to as our screening criteria were very few).
we prepared our questions in advance but did not take them with us and didn’t really follow a ‘script’ — not smart on our side. Next time, better keep a checklist of things to ask. And as with all types of qualitative interviews, the aim for the interviewer is to talk only 20% of the time and let the participant cover the remaining 80%. Practice will get us there!