Follow Datanami:
March 3, 2014

SAS and IBM King of Analytics Hill, But for How Long?

Alex Woodie

Last week’s release of Gartner’s very first Magic Quadrant for Advanced Analytics software put various vendors in a state of heightened excitement. It’s no surprise that SAS and IBM (via SPSS) are tops when it comes to making predictions from data. But a strong showing by smaller vendors–including RapidMiner, Knime, Alteryx, and Revolution Analytics–should keep the established vendors from getting too comfortable.

For much of the past decade, Gartner lumped analytics tools and business intelligence tools into the same Magic Quadrant report. And the analyst group still puts out an annual report, dubbed the Magic Quadrant for Business Intelligence and Analytics Platforms, that combines these two disciplines. In fact, it released an updated report for that product segment last week, too. (You can get your own copy courtesy of Tableau Software, which fared quite well in the report.)

But due to the growth in predictive analytics, Gideon’s firm decided to do things a little differently and release a separate Magic Quadrant focused entirely on so-called Advanced Analytics products. Gartner defines Advanced Analytics as “the analysis of all kinds of data using sophisticated quantitative methods [for example, statistics, descriptive and predictive data mining, simulation, and optimization] to produce insights that traditional approaches to business intelligence [BI]–such as query and reporting–are unlikely to discover.”

When it comes to “big data,” advanced analytic techniques like this are becoming increasingly necessary for the extraction of knowledge and meaning. It’s no longer enough to know what happened (descriptive) and why it happened (diagnostic), which traditional BI can tell us. Today, we want our tools to tell us what will happen next (predictive) and what we should do about it (prescriptive). You can read more about how Gartner sliced and diced analytics from BI in your very own copy of the new Magic Quadrant, courtesy of RapidMiner.

First, let us look at the advanced analytics vendors in the Leaders quadrant:

  • SAS dominated this field for years, with 40,000 customers in banking, insurance, business services, and government. The SAS product stack is the deepest and most complete, and its product quality and capability to model a wide variety of data sets is top-notch. SAS skills are widespread and easy to find. Besides the open source R stack, nothing comes close to the analytics giant from Cary, North Carolina. But SAS products are also very complex, with a steep learning curve (the upcoming SAS Visual Analytics should blunt these concerns). Cost is also a worry for SAS customers, and there were some product reliability concerns, according to Gartner.
  • IBM‘s 2009 acquisition of Chicago-based SPSS made it a leader in advanced analytics, and its Watson and ILOG products also give it some bite here. Gartner finds that IBM customers are happy with the overall quality of the software, including ease-of-use and speed of development. Other bright spots include the incorporation of entity analytics into SPSS Modeler and the incorporation of the Rapidly Adaptive Virtualization Engine (RAVE) with SPSS algorithms into Watson Analytics. On the flipside, all of these integration points for acquired products leaves unwanted complexity and create functionality gaps among the products from IBM and partners. Pricing is also a concern of users, especially when they must buy multiple products to accomplish one task.
  • Knime is a Zurich, Switzerland-based company that develops a popular data mining package that’s available via commercial and open source licenses. Gartner reports that Knime’s platform provides “an extensive breadth and depth of functionality,” and gave the company high marks for overall customer satisfaction. Ease of use and cost were also plusses for Knime, which occupied Gartner’s Leaders quadrant. The biggest concerns have to do with its low visibility outside of data mining, the “credibility” question about free and open source desktop software, and the difficulty that such a small company may have supporting customers.
  • RapidMiner was also placed in the Leaders quadrant, alongside Knime, SAS, and IBM. Formerly called Rapid-I, the Cambridge, Massachusetts-based commercial open source software company received high marks for product depth and breadth and overall customer satisfaction. The software, which can be extended with R and Weka, features templates that guide users through the most common predictive use cases. Like Knime, Gartner says RapidMiner struggles with credibility outside of data mining, and says the availability of free and open source software fails to motivate customers to upgrade to a paid version.

Executives with KNIME and RapidMiner were happy with their good showings. “KNIME is the only vendor in the advanced analytic platform space committed to the basic principle that a modern analytics platform simply must be open source,” KNIME CEO Michael Berthold said. “At the same time, we are adding layers of extra functionality that enable commercial customers to run the pure open-source platform more efficiently and more productively.”

“We believe our placement in the Leaders quadrant reflects not only our status as the world’s most frequently downloaded predictive analytics software, but also our ability to provide a platform that supports business analysts, data scientists, and business managers,” RapidMiner co-founder and CEO Ingo Mierswa said.

In the Visionaries quadrant are two vendors, Revolution Analytics and Alteryx, who also happen to be close business partners:

  • Revolution Analytics also fared quite well in Gartner’s Magic Quadrant. The Mountain View, California-based company is credited with realizing the power of open source R, and has therefore become “the default choice for organizations without an existing provider seeking an R-based solution.” High customer satisfaction and a strong sales pipeline are strengths. Weaknesses include the need for extensive R coding skills, unpredictable pricing (which the company addressed in late 2013), and concerns with product functionality in the areas of visualization and exploration/discovery, platform and project management and user experience.
  • Alteryx received high marks from Gartner for its approach to allowing analysts to “blend internal, third-party and cloud data, and then analyze it using spatial and predictive tools.” Built atop R, the drag-and-drop environment has high customer satisfaction ratings, and is often deployed alongside Revolution Analytics, which boosts Alteryx’s scalability (Revolution, in turn, benefits from Alteryx’s ease-of-use). There are questions about the Irvine, California company’s product reliability, and its ability to support the needs of true data scientists, and not just data analysts. And like Revolution Analytics, there are concerns about functionality in the areas of visualization and exploration/discovery, and platform and project management.

Dave Rich, CEO of Revolution Analytics, says he views his company’s rating “as validation of the innovation in the open-source R community, combined with Revolution Analytics’ enterprise expertise.” Meanwhile, Alteryx president and COO George Mathew says his company’s rating “affirms the need for a modern, user-centric platform for analytics. Legacy solutions continue to lag and aren’t addressing key areas where users need better tools to drive analytic decision making.”

Gartner put several other vendors into the Challengers quadrant, including StatSoft, a 30-year-old developer of data mining solutions based in Tulsa, Oklahoma; Angoss, a developer of predictive analytics software based in Toronto, Ontario; and SAP, the German ERP software giant that acquired BusinessObjects in 2007. Rounding out the report is the Niche Players quadrant, which was occupied by the likes of Actuate, Alpine Data Labs, InfoCentricity, FICO, Megaputer, Microsoft, and Oracle.

Related Items:

MapR’s Hadoop Offering the Strongest, Forrester Says

Big Data Future Hinges on Variety

Björk Was Wrong About Human Behavior, Big Data Says

 

Tags:
Do NOT follow this link or you will be banned from the site!
Share This