Inspiring reactions in just one screen

With clear goals and a dramatic presentation, dashboards drive conversations and build engagement around complex concepts.

The challenge:

Build a dynamic visualization that communicates at a glance how inaccurate and incomplete a customer’s data is compared to our verified, master data.

My team at Dun & Bradstreet built custom dashboards and reporting tools for sales engineers and account managers to visualize the power of cleansing and consolidating records against a master record set.

When you’re selling access to “Big Data”, there’s no product photography or app screenshots to build excitement around. You can’t just show a JSON blob and expect people to get excited.

Nor is it helpful to brag to a potential customer about the impeccable quality and breadth of your records; instead, you have to tell the story of how their data can be better. I designed and built a suite of data visualization tools for customer data; this is the story of the data quality profile dashboard.

Sketching out the data stories

Working with our solution architects who specialize in building custom data pipelines for customers, I broke down D&B‘s complex data cleansing and validation process into a series of stories to explain exactly how inaccurate and incomplete a customer’s data can be.

Each story became a small vignette, telling each chapter of a larger data narrative in a visually compelling and understandable way.

A stacked bar chart with the origin line at center makes it easy to see at a glance which fields are most often missing, wrong, or correct.
Pulling out specific examples of the worst data speaks to customers who have already noticed errors in their own records, confirming their suspicions that their data needs help.
Breaking out a pie chart into smaller sections shows multiple levels of detail in context.

Taken as a whole, these vignettes illustrated how D&B data could correct errors, deduplicate rows, and fill in missing fields — and what that meant for the customer’s records. We used them in various reports and dashboards across our suite of tools for data analysis.

Early concept for a data quality profile dashboard using one of the vignettes.
We built a full suite of data profile dashboards and tools that benefitted from the visualization stories.

Data quality at the 10,000-foot view

Once we had the basic idea for a series of visualizations, I came up with a series of high-fidelity mocks from which we could code our interactive data quality dashboard.

We chose this design to code up into an interactive dashboard.
More subtle variation of the prior design.

Once we picked the design, I built the dashboard in Vue, using animated SVGs and ChartJS graphs to display the data on the client at runtime.

A detailed data quality report

Part data profiler, part marketing brochure, the insights report was a custom-fit interactive experience.

There was far more to share about the customer’s data than just what the dashboard could display, so I built an in-depth reporting tool in the same style.

The online interactive report paired dynamic graphs with information about how to practically apply data insights from the visualizations.
Each chapter of the report provided a different detailed data analysis of a data story.

Like the dashboard, I developed this in Vue.JS with animated SVGs ChartJs. I also included a special print stylesheet so that customers could save and share a clean, well-formatted PDF of their report.


I make other stuff too!