Data visualization process demands smart design, accurate data
Article 1 of 4
The Charles Schwab Corp. has deployed Tableau's self-service BI and data visualization tools enterprise-wide to about 16,000 employees overall. The financial services firm uses the software to create dashboards for corporate executives, retail branch managers, financial consultants and other workers. There are even dashboards that track usage and performance data on the operational dashboards.
The analytics and data visualization process has become central to Charles Schwab's business operations, said Gessica Briggs-Sullivan, the San Francisco firm's Tableau administrator. The software itself is now part of the fabric that enables Schwab employees to do their jobs, she added: "It's not just a tool for many of them; it's a way of data life here."
As with many things in life, though, visualizing data isn't as easy as it looks, and doing it poorly can have big business consequences.
"If there are lies, damned lies and statistics, bad data visualizations are the viral means of spreading these lies," marketing analytics and reporting consultant Sri Chalasani wrote in a September 2017 blog post. Flawed visualizations may "lead to wrong conclusions that can hamper or -- even worse -- tank your business," said Chalasani, who works at technical services provider E-Nor.
She recommended a set of steps to help make the data visualization process effective. For example, designers should use appropriate charts, label elements correctly and avoid "cognitive overload" that makes it hard for business execs to discern and retain the information being presented.
Upfront data governance and curation are also needed to prevent users of tools like Tableau from distributing dashboards with embedded visualizations that are based on inaccurate info, said Rick Sherman, managing partner of consultancy Athena IT Solutions.
This handbook offers additional advice on how to manage data visualization efforts and get business benefits from them.