✍️ Emily Williams, June 2026

⏱️ 1-2 min read per step | 📚 20 linked resources (GOV.UK, NN/g, Google Sheets template + more)

I recently took informed consent in a lift. Don't worry, I talked them through the research an hour before and they still had the right to withdraw - I haven't joined the dark side.

I’d like to share 10 steps for guerrilla recruitment and naturalistic interviews based on what I learned from guidance I read beforehand and the fieldwork itself. You can scroll through, read the graphic in the Appendix, or jump to each step here:

In preparing for my transition from academic Psychology research to User Experience research, I recently kicked off a spare-time freelance project with Emma, a Java developer and website founder, and Iqra, a user-centred designer. Emma just launched a website that brings together all the tech events in Manchester, automatically pulling from Eventbrite, Meetup, LinkedIn, etc. I thought this was really cool and offered to lead some research.

We selected two events to attend and conduct some user research at, pending organiser approval. On Wednesday, I went to Silicon Mingle Manchester with Emma, which is a tech and start-up networking event (and not a speed dating event like my husband feared). On Thursday, Iqra and I went to Natter, which is a Manchester-based community for digital designers.

Work like this comes under a lot of names: guerrilla research, pop-up research, fieldwork, ethnography, contextual inquiry, naturalistic interviews, and intercept interviews, with some used interchangeably. I view the current work as guerrilla recruitment followed by naturalistic interviews (also pop-up / fieldwork). Here’s a field guide of 10 steps based on what I learnt.

1️⃣ The Plan 🎯 Land on a clear but flexible research plan, ideally with your team

No plan survives contact with the enemy

In field research, no plan may survive contact with the setting and participants, but it's still best to make one. As I recommended in my last post, addressing the "7 Deadly Sins of UX Research", pre-define a clear research plan. What is the problem statement, research goal, and/or research questions? Can you define these with your team? Who are your users, and how and where might you reach them? What topics do you want to discuss with them? (See Step 8 - The Questions). How many participants do you hope to recruit, from each user type? How long do you plan to speak with them for? Might you offer online interviews instead? Who will ask the questions, and who will take notes? Make sure your [team and stakeholders are all aligned](https://uxdesign.cc/tips-for-guerrilla-user-research-5fff52f30d0a#:~:text=Tip 1%3A Come to an agreement with your stakeholders.) (their Tip 1).

It's important to hold this plan lightly. Our plan was that I would conduct the interviews, taking brief notes, and Emma or Iqra would take more detailed notes. Our planned participant numbers, modes, and durations are shown in the excerpt below. When we were refining our interview questions and trying to select a smaller number for the shorter in-person interviews, we decided we would just try to ask our first in-person participant all of the questions and see how long it takes, gauging throughout whether their interest was waning. We would then decide which questions to ask in the shorter in-person interviews based on this first one.

Our plan to do only 2-3 short interviews in-person and only ask a reduced number of questions compared to online.

In reality, when Emma and I attended Silicon Mingle, we found that many more people were interested in sharing their thoughts than we expected, and the setting was especially conducive to seated interviews lasting around 20 minutes. After the first interview, we decided that we would just stick to asking all of the questions that we had planned for the longer online interviews. And based on the seated queue that had accumulated around us, we discussed whether we should split up and conduct interviews separately, sat at the same table. In the end, we interviewed 9 attendees and 2 organisers that day, lasting around 20 minutes each, in addition to signing up 1 attendee for an online interview the following week. Much more than the 4-6 attendees we had planned!

2️⃣ The Sessions 🗓️ Choose at least two and adapt your approach for each

Meet people where they are

Attending more than one session or setting is helpful if your users are likely to vary in things like age, user type, and work field. Attending Natter after Silicon Mingle allowed us to recruit more event organisers, some younger attendees, and digital professionals compared to founders and tech networkers. If you’re testing vehicle recall letters at carparks, two sessions at different times might help you reach more demographics. And apparently [Thursdays and Fridays are best](https://uxdesign.cc/tips-for-guerrilla-user-research-5fff52f30d0a#:~:text=Tip 10%3A If possible%2C do your research on Thursday or Friday.) (their Tip 10) for catching people at their most benevolent.

Compared to Silicon Mingle, Natter was much more structured than the description and "TBC" agenda suggested. Iqra and I had to take a different approach. At Silicon Mingle, we conducted mostly on-the-spot interviews. At Natter, the networking and break sections were 15-30 mins, not really allowing for our 20 minute interviews of the previous evening. I ended up just meeting people as normal, without a front-loaded research agenda. I felt that we had likely gotten enough brilliant input the previous evening, so it was no big deal if we didn't recruit anyone that evening. At the end of the event, I asked four of the people I had met if they would be interested in participating in the research, and I offered online interviews the following week to them. I also offered these to two of the organisers.

Event descriptions for Silicon Mingle and Natter.