Our research uncovered previously unknown pain points that lead to user dissatisfaction with the product. The insights and design recommendations will be considered and included in the product's next iteration.
As a team, we Iterated and delivered an improved onboarding process for an NLP product that enhances online shopping experience. As one of the four researchers on the team, I moderated 4 in-person user interviews, analyzed qualitative data, and provided actionable design recommendations. I was also responsible for creating the majority of the graphic documentations of our research process.
KLOA is a startup that focuses on making search easier by saving users typing effort. This new search experience consists of an input interface and a contextual search algorithm (see this section for interaction map). Ideally, businesses with mobile shopping apps like Amazon or Instacart would be potential clients of KLOA.
Previous diary studies on KLOA highlighted usability issues that the team is currently iterating on, including what types of feedback to provide users. This research then seeks to understand:
KLOA is aware of usability issues and is actively iterating on the design. So our study focused on two hypotheses:
While KLOA has a minimal design, the fact that the layout is challenging the status quo of QWERTY keyboard makes using it difficult for many. For KLOA UI's interaction details, check here.
While the interaction with KLOA consists with only a few steps (maybe even fewer compared to typing with QWERTY), the unique layout does make searching less intuitive for most users, at least at the beginning.
Per our client’s request and to better mimic the newest version of their design, we taped over the search bar and the word ribbon that existed in this iteration.
To understand users’ initial impression of KLOA and how new mental model can be constructed in a short amount of time, we designed a one-page instruction, introducing the basic functionality of the keyboard. Iterations of the instruction sheet are shown below.
All our sessions were conducted in libraries on campus. The participant and two members (one moderator, one observer) were in the room at each time. The rest of team observed on Zoom.
There are three components to our setup: camera 1, camera 2, and the prototype. Camera 1 captures participant's hand gesture and movement from above. Camera 2 captures participant's facial reaction and body language. The prototype records the user's interaction with the product.