Research
User Research Methods
Interviews
Accessing end-users for research was a challenge. I chose to go through the newly established Customer Success team because they frequently fielded support requests from admins at client organizations, and this information promised glimpses of what capabilities these power users wanted. I needed to know where they were getting stuck and what they wanted to achieve so that I could expand and support the skills they already had through the product's design.
I worked with my supervisor to identify which clients were most likely to grant us access to their end users. These clients were chosen on the basis of existing relationships, tenure with the product, and product usage. We included clients with both exceptionally high and unusually low usage to capture the product’s strengths and the areas in need of improvement.
We began recruiting and scheduling interview times with end users. Some of the calls would include user admins, while others would be strictly me, my supervisor as note-taker, and the participant. I wrote the interview questions.
User Research Findings
Personas
Note: Recent industry insights have revealed that personas may challenge efforts towards greater inclusion and equity in UX design. I no longer create personas; I've left this section visible only because it accurately depicts the process that was used at the time.
I created personas to represent the two types of users we met with. These personas reflect the insights gathered through ongoing relationships with our clients, with particular attention given to their ability and willingness to troubleshoot and their motivations for analyzing data.
User Journey Map
I created a user journey map to describe how end users felt as they answered their business questions using the product. A user journey map uniquely connects the user’s feelings to the product’s flow and affordances, and it helped me expand recognition of users' insight and skill throughout the organization.
User Flows
I created a user flow diagram to represent the current state of how a question is asked using the product. This helped illuminate where power users' skill is especially relevant, and where we can support expansion of an end user's query skills.
Next, I examined how this flow could be simplified. What are the essential steps in asking a question? Or in collaborating with a resident expert? Can we alleviate the end users’ need to seek support by simplifying the process of question asking? How do we best support power users as they answer local questions about the product?
Design
NOTE: My role with the company changed in July 2019, and I was not able to implement the design ideas presented here in the product. I chose to complete the design phase on my own at a later date, and the artifacts represented here are sourced from this independent study/practice.
Sketching
I leveraged the fastest, lowest-cost methods available to sketch a wide variety of ideas. I was exploring how we might encourage query constructure or provide support when needed.
One set of sketches focused on how a user could request support from their admin in the moment that they realize they would like assistance.
Another set focused on how to guide the user’s initial question construction.
Wireframing
Since the primary skill to support was query construction, I pursued a design that allowed the user to make adjustments to the questions they were asking in context. I created some wireframes to communicate the imagined flow.
Prototyping
This design was promising, but would need to be tested and strategized from a development perspective before we could pursue it. I created a high fidelity prototype for usability testing. This would hypothetically also be discussed with engineering teams to verify feasibility and estimate development time needs.
Testing
I wrote and conducted usability tests for the query construction flow. The results indicated that this design supported users in constructing queries, which accelerates finding insights. Unfortunately, we were not able to make the changes the test results supported.
Iteration
As natural language AI tools continue to evolve, there is ample opportunity to build on the insights discovered through this project. Contextual query construction - or prompt engineering, in 2024's vocabulary - is a burgeoning area of opportunity in many different professional disciplines, well beyond analytics.
Iteration upon the concepts here would likely apply these insights to different types of queries, or explore how else an interface can guide successful query construction.
Reflection
This project highlighted how technologies can expand users' capabilities. Research brought users' key abilities and pain points to light, and I used UX tools to discover how we could expand users' ability and relieve their frustration. I practiced stakeholder expectation management and end-user advocacy, skills I carry with me into any role or project I take on.