#9 User Research Can Be Fast & Easy
With a master's degree in psychology and many years of experience as a User Researcher, Nikki Anderson moved in 2018 from New York to Berlin. She works currently as a User Research Lead at Zalando as well as her own startup the User Research Academy.
Nikki shares the basics of qualitative user research to be able to get started with it in a fast and effective fashion. Next to that, she talks about when a company should start doing user research and at which point it's worth thinking of hiring a researcher.
Link to Nikki's Research Template:
Episode Minutes
Table of content
0:30 - Intro Nikki Anderson
2:46 - Researching the perfect croissant
6:55 - The basics of qualitative research
14:10 - Choosing the right segments to research
15:55 - Kicking off a research project
23:00 - The TEDW framework for user interviews
29:00 - Validating qualitative interview data
35:25 - Hypothesis-driven research (how might we statements)
37:35 - Evaluative research vs. generative research
41:55 - Parallelizing market research & user research
44:30 - When to start doing and hiring for research as a company
51:25 - Nikki’s key takeaways about getting started with research
54:00 - Debrief Christian & Alex
What makes a croissant perfect and how can we do research to figure that out?
Nikki: That's interesting. I mean, I'm vegan, so I'm sadly not an avid croissant eater, but I used to be! I used to be an avid croissant eater. I used to eat them with Nutella all the time straight from the jar.
For me, croissant research might be more similar to A/B testing than to actual qualitative user research. So I guess what I would do is I would start to ask people what makes a good croissant, what is it that you think about when you think of a good croissant? How would you describe this? And you might hear things like the flakiness or the butteriness or the crispness and all of those things. And then I think what I would do is I would go and work with my baker and A/B test, like a bunch of these different ingredients to see it until we got the right one.
Is this something you would do on a qualitative level or quantitative level?
Answer: Yeah, of course, you can always send out a survey to ask for data but there are a few things that happen with surveys, is that when you ask open-ended questions to get qualitative research insights, people don't like to fill them out!
I feel like the only people who really like to fill these things out are User Researchers because we know the pain that other User Researchers are going through to try and get this information. So, I would just go and talk to people and what you could do in this scenario, which would be really fun and I've totally done this before in New York but not in Germany yet... but you could go to a restaurant that sells these croissants and you could do “guerrilla research” where you go and you see people who are eating croissants. Then you interrupt their lovely time to ask them “Hey, What's up with this croissant?”, what do you think about this croissant? How would you describe it? How have you seen it done better?
That's what I would do and the reason I would do that instead of a survey is that you get rich data. So you get rich qualitative data from actually having conversations with people. That is super important if you're looking for insights because analytics and surveys can tell us what is happening.
But it's hard to get the why from that. So that's where qualitative research really comes in and compliments this data.
How do you run user research to evaluate a feature, problem, or idea?
The first one is qualitative research. We all know that we have low sample sizes. We don't have those crazy awesome sample sizes like we surveyed 1,500 people... We just don't have that because if I was supposed to talk to 1,500 people for one project, I'd still be talking to my first users at my first role. I wouldn't be here right now.
What I like to say is, and this is very common in the field, as you hear these, we need to talk to 5 to 7 users for usability testing, right? And then we need to talk to 10 to 12 users for more exploratory or discovery research. So there's some fine print that people don't like to talk about because it complicates things.
But what we need to point out is when I say talk to five to seven users, it's 5 to 7 users per segment! It's really sad to learn this because everybody's like, I can't just talk to five random people and be done with it. That's what I want to do. But 5 to 7 people per segment.
So what I mean is there are multiple ways that you can segment your customer base. If you have personas already at your organization those are a form of segmentation. So when you go to usability test something, you need to talk to 5 to 7 people within each persona that the feature would be relevant for.
Let’s say you have three personas and the feature is only really relevant for 2, you would talk to a total of 10 to 14 people because that's 5 to 7 per persona. If you don't have personas, you can segment in so many different ways. What I tell Researchers to do is look at different segments where your organization is making the most money.
Which framework do you use when you talk to users during an interview?
… tune in to learn more 🎧
