How do you apply the principles of human-centered design Agile (HCDAgile) to data projects? 1904labs employs our HCDAgile process for all of our projects, many of which focus on data engineering and decision science, where there isn’t necessarily a user interface. Knowing this was also a challenge many enterprises face, in October 2019 we brought together practitioners to discuss this very question.
HCDAgile for Data Overview
In this first HCDAgile for Data meetup, the discussion centered on two key points: what are approaches that leaders are taking in applying HCDAgile to data projects, and what are the challenges they face in doing so? With speakers from 1904labs and companies like ExxonMobil, the meetup was divided into several breakout sessions focusing on the research process, ways to visualize data, and overcoming objections.
This video features our HCD practice lead, Carol Righi, along with several Human-Centered Design Leads from 1904labs.
In data-focused projects, conducting research to understand the consumers of the data and their needs is critical to project success.
In this case, research involves involving other people in the process and gathering real user feedback. Conducting interviews, persona work, and journey mapping help bring along the technical teams who end up using the data system you’re working to develop - and they inform the solution.
In addition to gathering data, the act of conducting research allows team members to align through a shared experience of product discovery. While the research deliverables are personas, journey maps, etc., the other really important outcome is the experience of the technical team members. Their participation in this phase of the project helps them be part of the process to create the solution and builds buy-in and alignment to that solution, ultimately making it more sustainable, maintainable, and useful.
Often for data projects, the tangible user interface that results is some sort of data visualization, such as a dashboard. Everyone within an organization wants a dashboard to showcase the data they care about most. With that, it’s important to present data in ways that each persona will find meaningful, whether it’s for the CFO or an analyst. A dashboard can often feel like the outcome, but the true outcome is really what is done with the data in the dashboard and using that data correctly. The success is not just presenting data, but is the data solving a problem or facilitating an action?
Even if you’re not ultimately presenting that data in a dashboard but instead making it accessible via an API, those same principles still apply. What problem are we ultimately solving? How will that data ultimately be used? The answers to these questions help inform the way the data should be provided to be most useful to users.
Objections to the Process
One of the most common objections that practitioners face within an enterprise is taking the time upfront to study the current state. However, first understanding how data is being used today helps determine what should be done differently in the desired future state. Without this initial foundational understanding, a plan for change is just conjecture on what might work. By also understanding and documenting the current state, you then have a clear way to show whether your solution worked and improved that state.
One way to study current state is through value stream mapping. This is a lean technique to identify the flow of data through its completion, identify bottlenecks in that flow, and optimize it.
Our second HCDAgile for Data meetup will take place on April 30th at 12 noon CST. Register now for this virtual gathering of practitioners which will include a special presentation from Kate Walton, User Experience Group Design Lead at ExxonMobil.