Alexandra Wills is an ethnographer working at Fuse by Cardinal Health, an innovation center in Columbus, Ohio. She told this story on stage at Midwest UX 2017.

I’ll never forget when I did ethnographic research for a project aimed at helping a car manufacturer learn what Millennials with small children really needed.

The project was hard. Taking on a project at the height of the Great Recession meant navigating a radical change in client engagement from what I had experienced since starting the work two years prior. “It’s Friday at 5 p.m. in Ohio and you want me in Los Angeles on Monday?” Okay. “We’re doing video diaries and in-home interviews and a post-interview ideation session with participants in two cities, all in two months?” Okay.

Added to all that, I had a nine-month-old and simply didn’t want to leave her for days at a time. Over the past few months of work, I had already breast pumped on an airplane and in dirty airport bathrooms. I had already begged flight attendants and fast food workers for ice to put in the cooler carrying pouches of my “liquid gold.” Did I mention it was my birthday?

At one point in the project, I was hanging out with a family in Austin who had a toddler. I knew nothing about toddlers. After all, I had a nine-month-old. Did I mention I am not a ‘kid person’?

We had just returned from running errands in their car. As we got out of the car, they were showing me some specific details about the vehicle. They had a Honda Element – the car with the interesting doors that open and close like a book. I was paying close, close attention to the parents and I had no idea that the little kid was right near me. So I closed the door. Suddenly, we all heard the kid screaming! His parents rushed to his side and looked him over, examining his hands. All I could do was yell impulsively, “I didn’t do it!” I was horrified. I thought, “I hurt a child! This child! A participant’s child! Oh noooooo this is bad. How am I going to fix this? What am I going to tell Melinda (my boss)?” To this day I don’t know if his finger got caught in the door, or if me closing the door just scared him.

There was no blood, no broken fingers. But inside, I wanted to die. I already felt plagued by my own mommy guilt and that feeling spread throughout my body like lava. So, not only did I feel like a horrible mom for leaving my kid, but here I was in Austin, making someone else’s kid cry. What a moment. Needless to say, any rapport I had developed in my time with the family evaporated in that instant.

I stopped recording, stepped back, apologized to the mom and waited for the parents to finish calming down their kid. I waited for them to say, “This is over.” They didn’t. Miraculously, they continued the interview, even if I could feel all their judgment the entire time as we wrapped things up. “Maybe I didn’t traumatize this family,” I thought insecurely.

The icing on the cake was that we used video to capture all our data, so not only did this happen, but my boss got to see the whole thing when she reviewed the video. Later in the project I mentioned the incident and she said, “Yeah I saw that.”

Nadav Zohar is a UX researcher at AEP in Columbus, OH.

My first ever user research project was for a healthcare app. Our users were nurses who work with poor and high-risk patients, often called “the under-served.” My supervisor and I had a reserved conference room at the client’s site, and our pre-scheduled users rotated in about one per hour. It was a grueling two days of nonstop interviews. For the first day I took notes while my supervisor moderated.

On the second day, after he moderated the first couple of interviews, my supervisor turned to me and asked if I thought I was ready to take the lead on the next one. I said “Sure” so he handed me the discussion guide. In came our next user, a middle-aged nurse who was very sweet and eager to help us in any way she could. This was my very first user interview and I was ready for a clean, uneventful affair.

As the questions on my discussion guide turned to the technological hurdles she encounters when helping her patients, her frustration mounted. At one point, while discussing how her technology failed to help her manage the stress of the enormous workload placed on her and her colleagues, she mentioned having lost a patient. I watched her relive that pain – she broke down and started sobbing. None of the other users we’d talked to had even come close to that kind of emotional response, even though some of them had lost patients too.

Right then and there I learned there’s an awkward balance between not wanting to seem clinical and cold at that crucial moment, but still wanting to preserve an interviewee’s dignity: I figured weeping in front of strangers at work must be somewhat embarrassing. So I bowed my head and looked down at my notes, or my lap, or at nothing in particular, to give the crying nurse a bit of privacy. I waited a few sobs so it didn’t seem like I was trying to shut her up, and then I warmly and gracefully offered her a box of tissues. I let her know I empathized with her pain (although looking back on it I don’t see how I really could have…but my empathy felt genuine anyway) and she eventually calmed down and we finished out the interview. After that, back at the office I was jokingly known as the guy to call in to make people cry.

I think I deal fairly well with very emotional user research situations and over my career I’ve learned they are not uncommon, but it was interesting to have one right off the bat.