What I did and learned during my year-long UX research internship at ZapLabs.
From June 2016 to May 2017, I had the chance to work at ZapLabs as a UX Research intern for its business product, Zap. Zap is a CRM tool that helps real estate agent keep up with their clients, and gives them valuable insights into their clients' needs. Through working with different teams, I not only gained acute knowledge and flexible skills for working in an agile environment, but also met some great mentors and friends who inspired me to think outside the box when designing research studies.
Here's a quick glimpse of my journey at ZapLabs :
I interviewed the following users...
I answered these questions...
I applied these UX research methods...
UX research matters to all. To maximize the efficacy of UX research, it's crucial to keep product managers, designers and engineers engaged and updated throughout the research process. During my work, I constantly invited my colleagues to take notes or sit in during my interview sessions. Having them witness the users' responses at first hand and discussing with them about the key takeaways has really helped my research findings come through in the end.
Final deliverable is not the final chapter of a research study. To ensure that the research findings didn't just fall through the cracks, I found it important to follow up with product managers and designers to check what decisions was made based on the study, and what metrics were used to evaluate success of these decisions.
Be observant and initiative. For me, user research didn't only happen in interviews. I really appreciated that my team always sought for opportunities to play around with our own product. It helped me stay tuned with the latest release and self-diagnose issues that users might encounter. Having learned from my great mentor, Klara, I also tried to speak up whenever I had discovered or heard users talking about usability issues. Even though the issues might not be resolved immediately due to their complexity or other priorities, my colleagues actually took notes of them. If not, having an ongoing list of usability issues would continually remind people of the remaining issues.
Users may not know what they want. It sounded really cliche, I know, but it was something I need to regularly remind myself at work. Even though it's easy to say - we need to probe around why users want certain features and in what contexts they would use these features - users who have raised suggestions are not always able to articulate their goals or cases of usage well. In that case, I would need to go through a sequence of questions impromptuly to find out their real motivations. But this almost nit-picky interview process was totally worthy because later these knowledge of users' actual needs enabled me to give design recommendations that not only met their basic needs but also went beyond their expectations.
Take some time to get to know whom I am talking to. As an introvert, I found it difficult to make people open up quickly at the beginning of interviews. After asking my mentor for advice and reading articles online, I adopted the strategy of having some little chats with users before the session actually started to build rapport with my interviewees and allow users to drop their guard smoothly.
In real estate, a lead is a prospective property seller or buyer. When they request service on websites like Zillow or Redfin, sometimes their information will be sent to brokers. Before, brokers need to go to another system outside Zap to assign or re-assign leads to agents. Juggling two systems created lots of confusions for our users. The company decided to integrate the lead management functionality into Zap to streamline the process. However, we were lacking knowledge on user's workflow of managing leads.
In order to design a tool that met the needs of different types of users, we had to learn about how they managed leads in their daily workflow. I gathered insights from our target users, created a knowledge base for user personas and workflows, and delivered 4 design recommendations based on the findings from semi-structured interviews.
Method: Think-aloud protocols, Interviews
Duration: 6 weeks
My Role: Plan and conduct research, Analyze interview results, Present design recommendations
Since I didn't have any previous knowledge on lead management, I first talked to the product manager and another UX researcher who had worked on a related project to understand who were our target users for the new feature and what we already knew about them.
After our conversations, I also realized that the problem was more complicated than I initially thought. First, our target users did not have a consistent workflow. They belonged to three different real estate organizations which provided different tools for their members and had distinctive lead management rules. Thus, we need to interview each segment of users separately. Secondly, due to the complicated nature of our target users, I found that many of my colleagues were unclear about how the three segments of users overlapped and differed from each other. To keep everyone on the same page, I took the initiative to create a knowledge base including information on different company structures and user personas.
Collaborating with the product manager, I decided to focus the current research around the following questions:
In order to answer these questions, it was necessary for me to observe and inquiry about user's current workflows. Thus, I decided to conduct semi-structured interviews with brokers and company admins.
I conducted 11 remote, semi-structured interviews with users from two organizations (as we already conducted a related research project with the third organization). I started the interviews with warm-up questions, then asked my interviewees to share their screens and to show me what they do after logging into their lead management systems. I also probed around what they would do given four common lead management scenarios that I identified through the internal research.
Based on the research insights, I proposed that the new lead management function should:
I also suggested specific features that can help achieve these goals. To the date (01/18/2018), 3 out of 4 suggestions above have been implemented.