Microstrategy Provides Hadoop and Predictive Analytics with Version 9.3
July 27, 2012

MicroStrategy, announced version 9.3. The announcement came out of Amsterdam this month just in front of MicroStrategy World, the company’s annual conference for the European market. Release 9.3 delivers significant updates in four main areas: big data, advanced analytics, automated administration and visual data discovery.

The announcements on the big data front have to do with bringing data together from disparate sources, enriching available data, and new report search capabilities. Addressing the need to provide more automated support for data access and preparation are critical as found in our benchmark research on big data and our predictive analytics benchmark research as key obstacles to gaining business value from available data. The data source access improvements in 9.3 include improved access to departmental data, including data from spreadsheets and Salesforce.com, and from multidimensional sources such as Microsoft Analysis Services and Cognos TM1. The software can access data from SAP’s HANA appliance, and use a thrift connector to Hadoop distributions, including those of Cloudera and Amazon Web Services. The data enrichment enhancements include expansion of data based on ZIP code or date. Such location intelligence features address a hot area with great potential in the areas of database cleansing and enrichment. We’ll be exploring these trends in our upcoming benchmark research on location intelligence. MicroStrategy 9.3 also provides a Google-like function to discover reports and a dashboard, so users don’t have to spend unnecessary time looking for reports or creating new ones.

With respect to the Hadoop access, the company has four approaches. The first is to bring data from Hadoop into an in-memory structure for visual exploration and rapid prototyping using the imported data. This approach is interesting, but you still need to define your Hadoop queries before you do the analysis in memory, thus taking away the exploratory element of the big data. The second approach is to do freeform queries directly into Hadoop using Pig Latin or HiveQL.  This approach gives users back the exploratory aspect, but introduces complexity and sacrifices speed. The third approach is to model the data with a traditional multidimensional approach, while the fourth approach is to merge the Hadoop data with the enterprise data warehouse into a uniform view. The company says this last model is gaining traction for a number of their clients, which is in line with what we have been seeing from others in the space. Providing these options are critical as our big data benchmark research found that Hadoop as one of the key technologies planned in almost a third of organizations (32%).  

For advanced analytics, the new release integrates R statistical packages into the MicroStrategy BI platform, which allows for advanced in-database analytics with any available R algorithm, including many custom R developments. Version 9.3 supports the most-used algorithms straight out of the box with more than 300 functions. While others have integrated R, few have gone as far as integrating the visualization aspects of R, as MicroStrategy does in this release. R is well-known as an analytical tool, but most users don’t know about its visualization capabilities. The R language is gaining traction in both the academic and business worlds, with universities, large government organizations and the pharmaceutical industry all showing significant support. This integration of predictive analytics into business intelligence is an important step for MicroStrategy and our predictive analytics benchmark found 58 percent of organizations have this as a priority.

The third area of improvement is introduction of System Manager, a GUI administrative workflow tool that the company claims will reduce operating costs by more than 50 percent. The tool allows users to create administrative workflows from both MicroStrategy admin products and third-party tools to do things such as create an Amazon instance. Use cases include MicroStrategy intelligence reports, daily report execution schedules, and migrating objects. The package is priced separately, which is fine since this is a capability most BI packages do not offer.

The fourth and final area of improvement involves the already formidable Visual Insights, a visual data discovery tool MicroStrategy introduced last year. Visual discovery tools continue to gain traction in the market due to their ease of use and their ability to give time back to analysts. Our benchmark research into big data found that visualization is a top priority and unmet need in 37 percent of existing deployments just as is predictive analytics in 41 percent of organizations. The new capabilities of the 9.3 release include density maps, which help to highlight geographic concentration levels such as sales volume. Users can create network diagrams for analytics with web traffic, affinity marketing, or market-basket analysis, and image layouts, which allow for visual mashups. Other enhancements to Visual Insight include a wizard to suggest appropriate visualizations based on the data, the ability to do rank filtering, and shortcuts to commonly used metrics such as counts, moving averages and running totals.  Finally, the ease of creating and distributing dashboards is significantly improved. Drag-and-drop visualizations, and the ability to do visualization-to-visualization overlays, are impressive, and I expect to see others try to emulate these in the future.

Mobile Business Intelligence wasn’t addressed directly in the 9.3 release, but MicroStrategy’s platform for mobile applications was the focus of the 9.2.1m release in January. Mobile intelligence is a big part of the MicroStrategy strategy, and it was also a big part of the conference in Amsterdam. In a separate blog post, I wrote about Michael Saylor’s keynote speech, his new book, The Mobile Wave, and the company’s direction in mobile technology. MicroStrategy has been investing heavily in mobile for a while, especially around native support for Apple’s iOS.

In sum, the MicroStrategy 9.3 release is a big advancement for a firm already providing leadership in the analytics market. Given the firm’s advantage of being an enterprise platform and moving into discovery tools with Visual Insights, it is likely in a better position to expand than many of the discovery players trying to move upstream into an enterprise role. The fact that the company has built the platform from the ground up also gives it an advantage over some of the larger players with less than organic strategies. For organizations with MicroStrategy already installed, the 9.3 upgrade (and memory upgrades) makes plenty of sense. Any firm looking for deeper support of Hadoop, predictive analytics and visual discovery should examine this 9.3 release from MicroStrategy.

Regards,

Tony Cosentino – VP & Research Director


 

 Copyright © 2013 Ventana Research All Rights Reserved :: Privacy Statement :: Contact Us ::