Video: Just Build Me a Dashboard! (Videoconference)

100s of community videos are available to free members. Conference talks are generally available to Gold members.

Just Build Me a Dashboard! (Videoconference)

Tuesday, April 9, 2019 • Enterprise Community

Theresa Neil

Founder, Femovate by Guidea

Summary

How to bring data visualization best practices into your products, from dashboards to reports to BI tools. Learn about which charts to use when, creating accessible color palettes, and preview some new visualizations Theresa Neil has been testing in enterprise organizations.

Key Insights

•

Clients often request dashboards with vague goals, asking simply for something 'sexy' or flashy without clear user questions or business objectives.
•

Effective data visualization begins with understanding users and the specific questions they need answered, not just displaying all available data.
•

Many users, including enterprise professionals, struggle to accurately interpret even basic charts, highlighting the need to choose familiar and appropriate chart types.
•

False correlations frequently arise due to inappropriate use of color palettes borrowed from marketing, compromising data integrity.
•

Creating multiple tailored color palettes—qualitative, sequential, binary, and diverging—is essential for accessible and accurate data visualization.
•

Accessibility should harmonize with brand aesthetics; it's possible to design chart color palettes that complement brand colors while meeting accessibility standards.
•

Soft skills and diplomacy are critical when negotiating with marketing teams about color palette choices to avoid undermining data visualization quality.
•

Dark mode dashboards are highly effective in network and security operations centers where lighting conditions support such displays and improve visibility.
•

User engagement techniques like card sorting help prioritize the most relevant metrics, preventing dashboards from overwhelming users with excessive information.
•

Advanced experimental visualizations such as multigrain charts and correlated charts can provide richer analytical capabilities but require careful user testing.

Notable Quotes

"Make me a dashboard and make it sexy has been a common, but frustrating, request I've heard across industries."

"Starting with the users, not the data, is fundamental to creating dashboards that actually answer important questions."

"Many users, including smart enterprise professionals, are surprisingly poor at interpreting basic charts."

"False correlations often sneak into dashboards because inherited marketing color palettes are applied without adaptation."

"Color tells you where to look—our eyes are drawn to color first, so it must be intentional in data viz."

"We don't have to choose between beautiful and accessible; good chart color palettes can complement brand colors and still be accessible."

"You thank marketing for their style guide, identify non-negotiable brand colors, then build an accessible palette around those colors."

"Dark mode is particularly useful in network or security operations centers, where rooms are dark and displays must be highly visible."

"If you have 40 metrics, just putting them all on a dashboard doesn’t work; you have to prioritize through user research."

"In exploratory dashboards, layering charts and controls can facilitate deep data investigation; in explanatory dashboards, keep it simple and focused."

Previous video

Surprise me

Next video

Ask the Rosenbot