Jul 24, 2012 (10:07 AM EDT)
How To Choose 'Advanced' Data Visualization Tools
Read the Original Article at InformationWeek
There's data visualization and then there's "advanced" data visualization, but you'd be hard pressed to tell the difference based on press releases and marketing brochures.
Just about every business intelligence and analytics vendor out there has released an advanced data visualization module or add-on capability within the last year, with examples including IBM Cognos Insight, Microsoft Power View, MicroStrategy Visual Insight, new data-visualization capabilities bundled with the Oracle Exalytics appliance, SAS Visual Analytics, and SAP Visual Intelligence.
There's a rash of new tools because visual discovery is in big demand. IDC's latest BI and analytics market share stats, released earlier this month, show that Tableau Software, one of the leaders in advanced data visualization, was the fastest-growing vendor in BI in 2011 with a 94.2% increase in software revenue. Tibco Spotfire, another visualization leader, had 23.5% growth.
[ Want more on data visualization? Read SAS Introduces Big Data Visual Analytics Platform. ]
Visualization is hot because it makes data-analysis easier. Analysis with more conventional BI query and analysis tools isn't so easy, according to our 2012 InformationWeek Business Intelligence, Analytics and Information Management Survey. Nearly half (45%) of the 414 respondents to our poll, which was conducted late last year, cited "ease-of-use challenges with complex software/less-technically savvy employees" as the second-biggest barrier to adopting BI/analytics products. That was just behind the biggest barrier, "data quality problems," cited by 46% or respondents.
The online dating giant Match.com started using Tableau Software early this year because it wanted to put analysis capabilities "in the hands of our users, not elite analytics or BI experts," Atin Kulkarni, senior director of strategy and analytics at the Dallas-based company, recently told me.
Match.com's Tableau users now include product managers, finance managers, public relations people, and a group in charge of new business at Match.com--about two dozen users in all. Another 12 to 15 BI and analytics professionals and power uses at Match.com also use and appreciate the power of the software to illuminate patterns and trends that aren't as apparent when presented as data in columns and rows. Match.com is a Microsoft SQL Server Shop and it also uses SAS for advanced analytics, but it chose Tableau with more mainstream users in mind, according to Kulkarni. (Microsoft has since released Power View and SAS has since introduced Visual Analytics Explorer).
So what's the difference between "advanced" data visualization and routine sorts of charts and graphs you can do in Excel or Powerpoint? Kulkarni cites the example of geospatial visualization, something Match.com's new business unit is using to plan offline dating events. Match.com started putting on its own live events such as cooking classes and wine tasting parties as a way to extend its online dating business, but it requires a lot of analysis to plan an event that will draw Match.com users to a particular location.
With a data visualization superimposed on a map view, Match.com can see "where our members are located, what age group they fall into, and what gender they're hoping to meet," says Kulkarni. This helps planners see where concentrations of Match.com registered users and subscribers with the right chemistry can be brought together. With promising zip codes or neighborhoods identified, Match.com planners can then scour sites such as Yelp to spot popular locations to hold such an event. Visual reports can also show how many events the firm has already held in a particular area already, so a map of Manhattan, for example, would include pins representing events held in particular neighborhoods throughout New York City.
More than a few Tableau rivals would point out that they, too, can do map-based geospatial analyses. BI and analytics vendors are quickly upgrading data-visualization modules, adding this and that charting type to add depth to new products.
"It's an arms race where we say we have a tree map and a network graph and a bullet chart, and they'll say 'we have this one and that one, and you don't,'" says Lou Jordano, director of product marketing at Tibco Spotfire.
How do you separate the "advanced" visualization products from the also rans? In a new report, Forrester analysts Boris Evelson and Noel Yuhanna identify six traits that separate advanced data visualization from static graphs: dynamic data, visual querying, linked multi-dimensional visualization, animation, personalization, and actionable alerts. Dynamic data is the ability to update visualizations as data changes in sources such as databases. With visual querying you can change the query by selecting or clicking on a portion of the graph or chart (to drill down, for example). With multi-dimensional linking, selections made in one chart are reflected as you navigate into other charts. With personalization you can give power users an in-depth view and newbies a simpler view, and you can also control access to data based on user- and role-based access privileges. Visualizations can illuminate important trends and conditions, but what if you don't see the visualization? Alerting is there as a safeguard, so you can set thresholds and parameters that trigger messages whether you're interacting with reports or not.
Forrester's report, "The Forrester Wave: Advanced Data Visualization Platforms, Q3 2012," is available online from the SAS Web site. (The report was not sponsored upfront by any vendor, but SAS fared well in the research and purchased download rights for the report, as it often does with Gartner Magic Quadrant reports.) So that's what sets advanced products apart, but how do you pick the product that's right for your organization. Forrester's Wave report puts IBM, Information Builders, SAP, SAS, Tableau, Tibco, and Oracle in the advanced data visualization "leaders" wave. That's a pretty long list if you ask me, but the report includes a scorecard with individual 0 (weak) to 5 (strong) grades detailing more than 16 product attributes. Tableau, IBM, and SAP score highest on "geospatial integration," for example, whereas SAS, Tableau, and Tibco Spotfire score highest on visualization "animation," a technique used, for example, to show changes over time, in relationship to pricing changes, or other variables. Vendors in the "strong performers" wave include Microsoft, MicroStrategy, Actuate, QlikTech, SpagoBI, and Panorama Software.
I like Forrester Wave reports because the scoring and the weighting of the scores is spelled out in detail, so you can tweak the scoring formula to your own liking. For example, Forrester weighted 50% of its overall score of its assessment of current products and 50% on "Strategy." Within strategy, 40% of the score was based on "commitment" and 45% was based on "product direction" whereas only 10% was based on "pricing and licensing" and 5% on "transparency." Personally, I would make the strategy scores account for about 40% of the overall score, and I would raise the weighting of "pricing and licensing," as I'm guessing customers will care much more about that than "commitment," whatever that means.
One topic not covered at length in Forrester's report is the influence of big data. There's mention that insights into "deep and broad data sets" are easier to show with data visualization, and SAS and SAP are specifically cited for their big-data visualization capabilities. But there's no mention of the fact that vendors including SAS, SAP, Tableau, Tibco Spotfire and other have added connectors to Hadoop. Advanced analytic capabilities also get short shrift, mentioned as strengths for data visualization products from IBM, SAS, and Tibco Spotfire, but it's not treated as categorywide data-visualization attribute on Forrester's scorecard.
In my view, big data and analytics are crucial issues that are very much a part of the advanced data-visualization conversation. It's no coincidence all of the leaders of Forrester's Wave report have addressed big data, analytics, or both. Why? Because big-data insights and big-data predictions are more easily understood when they're presented in visual form. I saw this in action myself when I recently witnessed a demo of the SAS LASR Analytic Server and SAS Visual Analytics, the latter being the interface used to explore data on the LASR server. Drilling down on more than a billion rows of data, for example, I saw an analysis of six years' worth of manufacturing data with a predictive analysis of equipment reliability. Watch this video demo to get a better sense of the possibilities.