More Than Just Visualization
Like any enterprise product, the dashboard is one of the most requested features by users—partners, senior associates, project managers, and administrators. Each has their own goals and tasks that require different sets of data that can help them achieve their goals. The possibilities of what can be placed on the dashboard is endless, the question is — which ones will provide the most values to those that may benefit from them the most?
Role: Design / Research / Product
Collaborated with: Product manager / UX researcher / Developers
Kira has been a beloved tool among law firm associates and other contract review experts, as most of the earlier features were developed to do the heavy lifting of document review and analysis work. Now it’s time to ask ourselves — how would we help law firm partners and other senior people who may not be in the trench doing the work, but can equally benefit from the insight we can provide?
There are places in the app where both high-level and more granular information is provided, but due to a lack of cohesion in the way information is organized, it requires a bit of navigation in order to find useful information.
Having worked on a previous project that gave us some insights into the kind of information that was requested by each main persona, we decided to start the Dashboard project with partners, senior associates and project managers—those who care the most about high-level information but probably in different ways.
Planning for research
To understand the how a dashboard can serve their needs, we kicked off a round of descriptive research to validate a set of assumptions.
Since this is descriptive research, we produced a set of lo-fi mockups that captures some ideas we have heard from our users in the past and present it them to the research participants in order to gather more concrete feedback.
We transcribed research recordings into Dovetails, where we could code the transcripts flexibly.
After going through the codes and user quotes, we summarized them into a few themes, which we then translated to four main “how might we” design challenges.
It was quite clear at this point that the ability to provide clear, high-level, real-time information that gives partners or project managers a good sense of what’s in the project and what’s going on is the greatest value of the dashboard. However, the ability to drill down into details if needed is also important.
Displaying and visualizing the data is the easy part. The hard part is how we tell a story with the dashboard so it’s not just a pile of charts and eye candies.
In order to craft a story, we did a thorough mapping of datasets available in the product and how they address users’ needs.
Other than having a set of default visualizations that tell a story which helps partners and project managers make sense of their projects and status, it is important to make the visualization configurable. However, the overlaying of different facets of data poses a significant design challenge.
Audit of datasets
In order to make sure overlaying the different sets of data makes sense, we did a complete audit of all the data available.
Complex, but not complicated
Once auditing of the data sets are done, we sketched out the experience for the configuring and creating a chart.
To better the scope for the feature, we intentionally limited the chart type to bar chart, donut chart (which already exited in the app), and tabular.
Since we were also in the middle of auditing and refining the our design system, I also kept the colour palette of the chart to a minimum.
Designing a feature like dashboard, it is very easy to get distracted by fancy visualizations and lose sight of its primary goals and jobs to be done. It is important to always go back to the research learnings to keep track of those goals.
Be collaborative to make sure everyone is on the same page with feature scopes at all times.
Each product design project can be very different in terms of challenge and focus. In this case, auditing data structure, mapping how datasets can be combined, and designing for an intuitive configuration flow were the bigger challenge than coming up with the sets of visualizations to present.