You have your dashboard development tool of choice ready to go. Your data is ready. How do you make the story behind the data come to life for your target audience?
Tip 1: Determine Your Audience’s Needs
It sounds simple, but many dashboard projects falter because the driving needs were not identified before development began. Worse, these projects veer off course because the wrong set of needs were targeted based on faulty assumptions, miscommunications between users and developers, or some combination of the two.
In coaching budding analysts, I’ve often heard them express concern about asking too many questions. The truth of the matter is that most dashboards suffer because of a lack of domain knowledge. Experienced BI Developers start by interviewing their clients. They ask compelling questions: what questions are you trying to answer? What questions is your manager trying to answer? How well do you trust your data?
Before rushing straight into development, verify user feedback. Understand the hierarchies of user needs and the different levels of insights or details needed for different user groups. Deploy a functional requirements questionnaire. If possible, collect copies of their standard reports and ask what’s working and what’s not.
Tip 2: Play Detective
Delivering a compelling dashboard begins with curiosity. Deep dive into the data. Use the toolset of your choice to explore relationships between features, whether that’s Python, R, Excel, or the data visualization tool itself (Power BI, Qlik, Tableau). What you discover will help you to craft a set of data visualizations that display not only operational metrics but also illustrate relationships and previously hidden causative factors.
Design data visualizations to stimulate questions and conversation. A chart that shows how many widgets were purchased in the first quarter might fulfill your audience’s stated wish list, but it doesn’t contribute knowledge. A data visualization that illustrates contributing factors, capacity constraints, and recurring trends delivers insights and value. Perhaps the volume of widgets sold in the first quarter has been twice the volume of every other quarter, and your deep dive into the data discovers that two customers are driving 75% of that volume bump. That’s strategic knowledge the organization can use.
Tip 3: Make Sure Your Data Story Flows Logically
What distinguishes a well-crafted dashboard from a collection of charts? A well-crafted dashboard tells the story without requiring explanation. Each chart is a story point, and the story points flow in a logical sequence.
The most visually compelling dashboards present their highest-level insights where the user is accustomed to start reading. In Western cultures, that starting point is in the top-left quadrant of the page. The data story then flows across to the right. More detailed insights are positioned in the lower half of the page.
Tip 4: Use Colors Strategically
Your users are looking for insights quickly. Help them out by directing their attention to what’s important. In this example, projects that are “stuck” are highlighted, making users aware of projects that are in need of attention or resources that require support.
Traffic light colors can also be effective, but there are great alternatives for users with color perception deficiencies. Explore alternative color schemes to support 508 compliance for projects with large audiences.
Tip 5: Less is More
Dashboard development tools (Power BI, Qlik Sense, Tableau) enable you to design data visualizations and dashboards more quickly than ever before. But beware of “standard” chart features that distract from your data story.
Some dashboard development tools automatically add chart features that may be less than optimal, such as axis lines or legends that clutter the visualization. Your users will appreciate exact numbers or dimensions on data visualizations, instead of taking extra time to interpret. As a best practice, remove distracting axis lines or legends and use data labels to specify metrics instead.
There you have it – five tips to creating dashboards that aren’t just pretty pictures, but effective and insightful analytics support for your organization.
Data Scientist at Definitive Logic