In-Depth

Tableau Preps Next-Gen Data Viz Tool

Orgs are increasingly hip to the importance of data visualization, especially as an enabling technology for analysis, Tableau sez.

• By Stephen Swoyer
• 06/07/2006

When Tableau Software released version 1.5 of its eponymous business intelligence (BI) suite last October, its bread-and-butter technology segment—data visualization—hadn’t yet gone mainstream.

Eight months later, data viz still hasn’t gone mainstream, but it has picked up a few big name converts, including Business Objects SA (which last November acquired visualization specialist Excelsius), Hyperion Solutions Corp. (which notched an OEM agreement with Tableau), and Microsoft Corp. (which came into some first-class visualization capabilities of its own when it acquired the former ProClarity Corp. several months ago.)

Next week, Tableau plans to unveil a version 2.0 release of its flagship visualization software, chock full of new capabilities, including support for relational database joins, offline access features, and improved analysis and statistical modeling capabilities.

Tableau chief Christian Chabot says it’s as good a time as any to be in the data visualization biz. He cites Tableau’s strong quarter-to-quarter growth and big-name customer wins—including companies such as General Motors Corp. and Safeway Inc.—and says that Global 2000 organizations are increasingly getting hip to the importance of data visualization, especially as an enabling technology for analysis. In most cases, Chabot claims, customers tap Tableau to help complement their existing BI front-end tools.

“We often sell to people who have every package under the sun, literally. Safeway has Hyperion, BO, stuff from Teradata, and they have homegrown systems, they have SAS, and they’re one of our biggest customers,” he comments. “In those types of sales, we are usually serving the role of the specialty upgrade for the intense visual analysis capabilities that we have.”

But Chabot downplays the suggestion that visualization itself is a specialty niche. Instead, he says visualization is a means to help bring BI down from the mountain, so to speak. “This whole industry sits around all day talking about bringing BI to the masses. But which product is doing that? There is no such product. Everyone thought it would be [Microsoft’s] Business Scorecard Manager, but frankly BSM and all [of] these [products] take weeks to deploy and cost a bloody fortune,” he comments. “What you really have to have to do BI for the masses is number one, [the product] would have to deploy in seconds. Unless you’re talking about a Visio-style install, it ain’t BI for the masses.”

Tableau isn’t quite turn-key, Chabot allows, but it does boast canned connectivity to a range of different data sources—including relational, multi-dimensional, flat file, and certain other kinds of repositories. More to the point, he says, Tableau doesn’t just rely on generic connectivity into target data sources. To get at Hyperion’s Essbase OLAP engine, for example, Tableau uses that company’s flavor of MDX. Ditto for SQL Server Analysis Services. And to get at relational data, Tableau doesn’t rely on ODBC connectivity, either, but exploits native relational database interfaces when and where possible.

“What we do is we work with each database vendor individually to connect to their highest performance and internally recommended API,” he says. “So we didn’t just go expose a generic ODBC connector, because just having a lowest-common-denominator approach, our vision wouldn’t work.”

As a result, Chabot claims, Tableau can connect quickly and reliably to target data sources. This helps speed time-to-deployment. “We have customers telling us, ‘In our first day, we literally can see or understand [our operations] like we’ve never seen it before.’ We’re presenting graphics of what’s actually going on that they’ve never seen before,” he comments.

The Tableau 2.0 release his company plans to unveil next week will augment these capabilities, Chabot says. “We’ve added dynamic calculations [such as the ability to do] table calculations, moving averages, compound growth rates, cumulative totals—a whole class of calculations Tableau did not have before,” he indicates. “And with our offline [analysis] capabilities, we make it easier for [an analyst] to connect to a mothership relational database and do their analysis of it [offline] and then come back online [to update].”

Other Tableau 2.0 enhancements include support for trend lines and statistical modeling (for power users); support for large database analysis (in which users can analyze a random sample and then run it on the entire data store); data playback capabilities (such as paging and inspection and data animation); as well as improved exploratory modeling capabilities (with support for dynamic trend lines, visual modeling, and forecasting).

Speaking of BI for the masses, Microsoft this week announced an ambitious expansion of its already ambitious 2007 Office System BI strategy, but Chabot says Tableau is a great complement for next-gen Office. “Excel is obviously a fantastic vehicle to get BI to the masses. It’s relatively easy to use, it’s relatively easy to deploy, and it’s relatively easy to get access to,” he concludes. “We’re excited about that. We bearhug it and upgrade it and integrate with it extremely well. So we can take you from Excel-level BI up to Cognos-level BI, while again embracing the [ease-of-use] model of downloading and deploying in minutes.”