4 Lessons I learned conducting user interviews in Germany
Last night, I returned from 4 weeks of conducting user interviews for a product that will be used in Germany. This means, we needed to understand native German users. For the record, I do not speak German. And actually no one on this team was a native German speaker. So when it came time to conduct user research, we made a few mistakes. I wanted to recap some of what we learned here
1. Avoid a translator
We learned this the hard way. We assumed that our users would be too uncomfortable speaking in English so we brought in an outside user research team to help out. We briefed the outside team on our product, user needs, goals, and even gave them a research script. They were responsible for conducting the interviews, translating the highlights, and sharing those highlights with the product team.
By assuming that our users would only be able to communicate in their native German, we got two things very wrong:
- We attributed too much of the research findings to the words that interviewees use during a user interview. That’s not a fair assumption. We can actually gain more insight from how they interact with stimulus than we can from the exact words they use to express themselves. This is true across native language interviews, too.
- We assumed that interviewers didn’t speak enough English for us to gain any value. But when we started conducting interviews in English, even among less educated communities, we found that the language requirements for interviews were pretty minimal. We were talking to users about their favorite hobby, and it’s often a person’s hobby that they can discuss in other languages because they’re often part of an international hobby community. (Like discussing your favorite soccer team.)
- We didn’t want to make the interviewees uncomfortable because we feared they might not be able to share as much. It’s true that our interviewees were a bit uncomfortable in English. But as the hour went on, they became more and more comfortable speaking with us. And, this assumption ignores the fact that most user interviews are uncomfortable for the user anyway (across all languages). Yes — maybe our users were a bit more uncomfortable. But a good research team is almost always navigating an uncomfortable interview. It’s a critical skill as a user research team to be able to make users feel comfortable chatting about lots of different topics.
Beyond the wrong assumptions, the translation process was a huge resource/time waste. We spent way too much time briefing an outside team to do interviews, translate the interviews, and synthesize them. Instead of spending 1–2 days per week, it controlled our entire week of work. Once we started asking people to interview in English, we seriously cut down the resources it was taking to do interviews.
Of course, if you have the ability to conduct the research in a foreign language that the whole product team understands, than definitely go for it. But that was not our situation. 75% of our product team was far from fluent speakers.
2. Bring visual queues
By the time we started asking for interview candidates who spoke some english, we already had some wireframes that we could put in front of them. We structured our interview so we asked them some general questions before presenting wireframes, then presented paper wireframes / prototype. The wireframes gave them something to hold, interact with, and reflect on.
Unlike informational interviews, understanding how a user interacts with wireframes requires emotional and visual listening. You get to watch what they do and compare that to what they say and understand. In this case, the words the user might say are actually secondary. For example, one user kept comparing our wireframes to a restaurant finder — that was not what we were trying to communicate — but by watching how he interacted with the wireframes, we could see that he understood the real goal of the wires, which was searching in based on relative location. His words said “restaurant” but his behavior actually communicated understanding.
3. Give the user a marker
We set up our interview room with big pieces of paper, colored markers, and thick pens. Our users were a bit nervous about picking up a pen, at first, but once we got them writing, some of them didn’t want to leave. One of our most excited users stayed an extra ten minutes to draw on the wireframes and show additional features he thought would benefit the project. He might not have been able to verbalize his creative ideas in another language, but the marker and paper allowed him to truly show us what he understood about the project and where he thought it was going.
By drawing or writing down words in native German, we gave them another means to communicate that lowered the language barrier.
4. Recruit the users yourself
We struggled with this one too. We weren’t familiar with the local community of users so we enlisted a third party agency to help us find relevant users. They weren’t familiar with our nuanced user needs andy they weren’t comfortable providing different types of users on short notice (in one week we needed to stop taking to 30 year olds and start talking to 19 year olds and that was really hard for the agency to accommodate. )
We used the agency because we were unfamiliar with the culture and concerned about the language. But as an experienced product team, we shouldn’t have let this get in our way. It’s our job to get to know new communities and their nuanced need. I don’t know any chemistry but it wouldn’t stop me from working on a pharmecutical app. There is always jargon/language/cultural differences that product teams need to overcome. Why did we let a simple foreign language get in our way?
After 3 weeks of unsuccessful third party recruiting, we took on the recruiting ourselves. We utilized all the skills we normally use when taking on a new vertical and found the communities we needed to reach. Within one day, we had a list of participants that would speak english and we could qualify them ourselves if they were close to our persona based on their public Facebook profiles.
A language barrier was so intimidating that we lost some faith in our daily skill set — but in reality we could have used our standard tools from the outset.
