TDWI | Training & Research | Business Intelligence, Analytics, Big Data, Data Warehousing

The Dashboard Reloaded: Tableau Touts Version 3 point 0 Release

Tableau 3.0 delivers a host of niceties, including support for rich formatting, free-form annotations, and—chiefly—a new Dynamic Dashboards component.

• By Stephen Swoyer
• April 25, 2007

The data visualization space isn't exactly bereft of competition, but Tableau Software Inc., along with visualization stalwarts Advizor Solutions Inc. and Spotfire Inc., seems more competitive than most.

Last week, Tableau unveiled a version 3.0 release of its eponymous data viz offering. The revamped Tableau 3.0 delivers a host of niceties, including support for rich formatting, free-form annotations, and—chiefly—a new Dynamic Dashboards component.

Tableau officials say the new dashboard feature—which supports simultaneous connectivity into heterogeneous data sources—isn't just a rehash of the static dashboards of old, or Dashboards 1.0. What else could you expect from a data viz specialist, after all?

"Previously, people would do investigation and analysis; they'd create these worksheets just like they would in Excel, and they kept asking us for the ability to aggregate these things into something they could basically use as a monitoring tool, i.e., a dashboard," says Kevin Brown, vice-president of marketing with Tableau. "We said to ourselves, lots of dashboards exist in the world—how can we make this better? So we took the dashboard concept and made it fully Tableau-interactive. So you can create a dashboard from multiple data sources. One panel might be connected to an Oracle data warehouse, one to Excel, one to SQL Server, and they all refresh dynamically. That's a pretty big deal to have a heterogeneous interactive dashboard."

Brown contrasts this vision with many of the so-called Dashboard 2.0 implementations that exist today. True, such dashboards do frequently refresh, but many don't support dynamic drill-down, or other user interactive features.

"A lot of these [dashboards] are based on some specifications that someone gathered, where they've defined specifications, configured [the dashboard], and implemented [the dashboard]. By and large, it would mostly be static. There would be some capabilities to pivot the view or change the date range, or maybe to filter, but—by and large—if you've ever worked with corporate BI dashboards, they tend to be fairly static, and they're not changeable," Brown argues. "So if you have a different view or a different idea, you have to go to the BI guys and the IT guys to try to convince them to configure it differently."

Enter Tableau 3.0, which Brown says makes it easier for users to build drag-and-drop interactive dashboards—in many cases without intervention from IT.

"We give the user the ability to take your Tableau analyses or reports and basically through a drag-and-drop interface configure your own dashboard in minutes and have it fully interactive," he says.

The trick, Brown explains, is that users can drill through the dashboard right on down to the underlying data, so Tableau 3.0 is actually able to service a range of different user classes—from decision-makers who want at-a-glance information access, to business analysts or power users who need to peer ever deeper into the data.

"In order to create the dashboard, I just pick and choose those [elements] that I want to juxtapose. You can zoom these things, change their views. The idea is that a customer might have four or five of these dashboards that are available to whoever they want to make them available to. Instead of having the IT guys or the BI guys having to configure these, it's as simple as dragging and dropping these views into a dashboard," says Brown.

In this respect, Brown contrasts Tableau's new Dynamic Dashboard capability with those of other dashboard purveyors. "As opposed to monitoring and visual reporting, this is more of an analytical dashboard. It's sort of a different style of dashboard with a bigger, more interactive feature set. It's something more than just a dumb, static monitoring product," he argues.

Tableau 3.0 also boasts advanced filtering capabilities, according to Brown—including support for dynamic update features. This means users can create and adjust a filter such that an array of visual displays update simultaneously.

"If you're looking at a view of data that has a hierarchy that's different from the way you want to look at the data set, you've got two options, either change the underlying hierarchy or you can use Tableau," he explains. "Tableau doesn't change the underlying data set, [it] does all of the aggregation, all of the mapping for you automatically. This is really a classic problem in BI, [where] the data you want to look at is too detailed. It's basically creating ad hoc groups, we call them ad hoc groups, or binning and categorizing."

Another new amenity is an annotation feature, which lets users add free-form comments and dynamic data labels to Tableau visual displays. The idea, Brown says, is that such annotations are dynamic and are automatically repositioned by as a user continues to explore and analyze. "This whole idea of doing visual analysis and letting the user annotate, in Tableau 3.0, we've got free-form analysis. In other tools, you'd have to do your analysis and then move it out to PowerPoint to do a bunch of beautification," he argues. "What you get are very nice, professional, polished views that are annotated inside Tableau."

Nor is that all. Tableau 3.0 also features support for rich formatting. What this means, Brown explains, is that users can build presentation-ready tables that maintain their live connection to underlying data sources. Table formatting gives users control of fonts, colors, and shading on all axes, rows, and columns, Brown says. "A lot of people like text summaries. Reports are really useful. And now you can create these nice, rich, formatted reports, too."