Adapted from a lightning talk presented at MERL Tech 2018 in Washington DC.
As a data visualization developer, whether I’m at a conference or speaking with a new client, the most common question I hear is “What dashboard tool should I use?”.
I’m not surprised; technology platforms make big promises. Reading the teasers on a quick web search, Tableau promises to “help anyone see and understand their data” and allow users to “share with a click.” Qlik tells users that their dashboards will “instantly pivot your thinking.” And Power BI offers to “Provide a 360-degree view for business users…updated in real time.”
While each of these technology products (and many others in the market) offer a range of amazing features, when we’re so focused on the technology we can lose sight of our ultimate goal: to empower people with information they can use. We become so focused on picking the right technology, we often forget that the dashboard we build is a digital product in itself.
If we think of our dashboards as digital products designed to serve their users, we can look to design to consider what makes a successful dashboard. Consider if the dashboard is:
- Desirable: Does the end user find the dashboard to be intuitive, interesting, and insightful for answering their questions?
- Feasible: Can you build the dashboard with the resources – people, platforms, and data – that you have access to, or can procure?
- Viable: Can the dashboard refresh and automate (or something close) in order to minimize recurrent tasks and time required for updates?
Missing the Mark (and Common Issues We Can Address)
When dashboards are launched but not used, we’ve often skipped the first criteria above: desirability. By refocusing on user needs first (rather than tech stacks), we can deliver dashboards with which users are excited to interact.
In nearly a decade of developing data visualizations across the public and private sectors, I’ve observed some common user issues that can be easily addressed by giving greater consideration to user needs, product usability, and launch.
“I don’t know how to use the new dashboard.”
- Often rooted in confusing UX design or ineffective or insufficient training for new users.
- Ideas: Ensure user cues are embedded in the dashboard and provide on-demand trainings. Don’t assume a user will go looking for a PDF user guide saved in a shared folder.
“I can’t find the data I need.”
- Could be a result of user needs not being well understood or prioritized from the start (and the data really isn’t available in the dashboard) or a failure of design and usability if the data is buried behind a set of dense views and filters.
- Ideas: Co-design with end users from the start, and ensure they’re partners in the design process. Create multiple points of feedback. Make purposeful design decisions to address user needs, and make sure you conduct user testing. Observational sessions where a developer listens as a user narrates their experience (and identifies where they get stuck) are feedback gold!
“The dashboard doesn’t help me uncover why something happened.”
- When design doesn’t enable appropriate levels of exploration and drill down (poor information architecture) or fails to incorporate necessary context, users are left with a snapshot without comparisons or benchmarks.
- Ideas: Link views within dashboards and enable drill downs (often a built-in feature for dashboard tools if you have the necessary granularity in your data). Add reference lines, averages, and benchmarks for quick visual comparisons and on-demand access to the context your user needs to interpret the information.
Dashboard platforms can enable great design but aren’t silver bullet solutions that can solve all of an organization’s data challenges. By considering user needs first, we can develop better dashboards that users are excited to log on and use.