Blog: Wayne Eckerson Subscribe to this blog's RSS feed!

Wayne Eckerson

Welcome to Wayne's World, my blog that illuminates the latest thinking about how to deliver insights from business data and celebrates out-of-the-box thinkers and doers in the business intelligence (BI), performance management and data warehousing (DW) fields. Tune in here if you want to keep abreast of the latest trends, techniques, and technologies in this dynamic industry.

About the author >

Wayne has been a thought leader in the business intelligence field since the early 1990s. He has conducted numerous research studies and is a noted speaker, blogger, and consultant. He is the author of two widely read books: Performance Dashboards: Measuring, Monitoring, and Managing Your Business (2005, 2010) and The Secrets of Analytical Leaders: Insights from Information Insiders (2012).

Wayne is founder and principal consultant at Eckerson Group,a research and consulting company focused on business intelligence, analytics and big data.

January 2014 Archives

While a number of established BI vendors are showing signs of duress due to the market shift from reporting to analytics and the onslaught of sexy, new visual discovery products, one old-time BI vendor that is bucking the trend is Datawatch, a public company whose stock price has skyrocketed from 11 to 33 in the past nine months.

Founded in 1985, Datawatch has toiled out of the limelight for many years, selling decidedly unsexy "report mining" tools, which enable customers to extract and integrate data from print spools, ASCII files, EDI feeds, XBRL and PDF documents and just about any other formatted document or report as well as relational sources and output the results to BI tools, Excel, HTML, and other interactive environments.

Along with its low profile, Datawatch had flat revenues. The problem, according to president and CEO Michael Morrison, who was brought in three years ago to turn around the company, is that Datawatch left a lot of money on the table because it didn't offer a front-end BI or visualization tool. A former IBM Cognos executive, Morrison moved quickly to restructure the company's operations and hire industry veterans from Cognos, Applix, Hyperion and other leading BI vendors to rewrite the company's strategy and execute it.

The move that catapulted Datawatch's stock this summer was its acquisition of Panopticon, a real-time visualization vendor founded by a Swedish banker that targeted capital markets. Investors quickly recognized that in the era of big data and visualization, the combination of Datawatch's report mining technology--which it now calls "information optimization"--and Panapoticon's real-time visual discovery tools make a potent combination.

Datawatch now bundles both technologies in its Datawatch Desktop and Server products, which compete directly against Tableau, Qliktech, Tibco Spotfire, and other vendors in the red-hot visual discovery market. Unlike those vendors, Datawatch handles streaming and multi-structured data, giving it a decisive advantage in many customer accounts. It has also sealed partnerships with Streambase, Splunk and other data management vendors to deliver real-time streaming analytics against multi-structured data.

With high expectations from investors and others, Datawatch is now the one to watch. Datawatch is based in Chelmsford, MA with stock symbol DWCH. For more information, go to: www.datawatch.com.


Posted January 27, 2014 4:10 PM
Permalink | No Comments |

Big data, by itself, is useless. Think about it. Data is simply a stream of 1s and 0s. It's not until you analyze data or use it to support or automate a business function that you gain any value.

To date, the big data movement has focused on technology and components and how to make them easy to use. The big software and systems providers--IBM, Oracle, SAP, Pivotal (EMC/VMWare), Hewlett Packard, Microsoft, and Teradata--have all recognized the changing tide and raced to acquire, assemble, and integrate various big data technology and components. But the next wave of competition is delivering real business value with these newly architected big data platforms.

One small company that aims to leapfrog the competition in this way is Infochimps, a big data startup which last summer was acquired by CSC and is now part of its 1,000-person "Big Data and Analytics" business unit run by veteran Andy Walker. CSC is betting on the Infochimp Big Data platform to help it deliver faster, better, cheaper big data solutions to its clients.

Given that strategy, CSC has repositioned Infochimps as a cloud service that offers three independent yet integrated services: 1) real-time streaming for integration and analytics applications 2) a NoSQL service for ad hoc analytics and 3) a Hadoop service for batch analytics--all of which will operate on public, private, and virtual private clouds.

In addition, CSC Infochimps is applying CSC's domain knowledge to create packaged applications running on Infochimps for media, retail, manufacturing, financial services, healthcare, and telecommunications. The cloud interface hides the complexity of all the open source (e.g. Storm, Kafka, Hbase, ElasticSearch, impala, Stinger, NoSQL, Hadoop, Hive, etc.) and commercial components required to deliver a complete solution.

So the little startup with the funny name could make a splash if CSC can turn it into a packaged solution rather than simply a tool for selling complex and costly consulting services. Stay tuned!

For more information, see www.infochimps.com.


Posted January 22, 2014 1:33 PM
Permalink | No Comments |

What is a sexy statistician? A data scientist! That is, of course, until the statistician has to learn how to write Java and parallel code to unearth patterns and outliers in "big data" a.k.a. Hadoop. Then, the fledgling data scientist reverts back to homely statistician.

Fortunately, startups are rushing to fill the void with big data tools for statisticians that enable them to skip the coding and focus on what they do best: massage and model data. One of the recent newcomers is Oxdata whose H2O prediction engine supports the complete analytical workflow, from preparing to modeling to scoring data. Its secret ingredients are an in-memory distributed key value store (think memcache) and super-fast distributed algorithms and scoring routines, including classification, regressions, clustering, and ensemble modeling.

Moreover, H2O is open source software, so you can download the platform, console, and algorithms to your laptop free of charge. H2O requires four times the amount of memory on your machine as data you want to process, so "big data" is only as big as your memory footprint. But you can import CSV, XLS, Hive, and ARRF files, among others. Once downloaded, you run the software via a browser, an R Console or RStudio, Python, or any application via a REST/JSON API. Several large insurance companies are using H2O as well as Netflix and Trulia.

For more information: For more information, go to www.oxdata.com.

Companies with similar products: Alpine Data Labs, Predixion Software, Skytree, Revolution Analytics, Rapid Insight, SAS Visual Analytics, SAP Predictive Analytics, Oracle Predictive Analytics, IBM Predictive Analytics


Posted January 21, 2014 1:17 PM
Permalink | No Comments |

Through a recent consulting assignment, I discovered that there are many Java-based reporting tools on the market, each carving out a profitable niche in the industry, generally out of the spotlight.

One of the more promising Java-based reporting vendors I've met is Jinfonet Software. Founded in 1998, Jinfonet, until recently, sold its flagship product, JReport, as embedded software to other software vendors. Today, Jinfonet is taking JReport upstream, selling it more aggressively as a stand-alone product with new modules for dashboarding, self-service reporting, visual analysis, Web and mobile delivery. It also provides access to relational, NoSQL, Hive, and Amazon Redshift data sources, and it's working on modules for predictive analytics and collaboration. JReport is no longer just a Java-reporting tool, it's a full-featured BI platform.

Jinfonet has 1,000 customers, 60% of which are ISVs, but a growing number are user organizations. With its CPU-based pricing, many of these customers, such as Visa and HSBC, use the product to support large extranet deployments. (Jinfonet also offers named user pricing.) And unlike some of its Java-reporting brethren, Jinfonet has modern-looking graphics and is eager to get the word out about its capabilities. If you are in the market for a full-featured Java reporting tool that is keeping pace with the world of analytics and big data, then Jinfonet is a good place to start.

Similar vendors: Jaspersoft, LogiAnalytics, Actuate, Yellowfin, Pentaho, Inetsoft, Quadbase, Winward

Jinfonet is based in Rockville, MD. For more information, go to www.jinfonet.com or call (240) 477-1000.


Posted January 21, 2014 12:13 PM
Permalink | No Comments |

Visual discovery tools have breathed new life into the business intelligence (BI) market. They are not only fun to use, quick to deploy, and easy on the eyes, they are the fastest growing segment of the BI market, thanks to pioneers QlikTech, Tableau, and Spotfire (now owned by Tibco), whose growth has trimmed the market share of traditional enterprise BI players. In response, most other BI vendors have recently introduced their own visual analysis products to widen their product capabilities and protect their installed bases.

Birst is the latest BI vendor to enter the visual discovery fray with its Birst Visualizer tool. While most enterprise BI players have simply shipped Tableau look-alike products whose only value-add comes from offering the software free of charge to enterprise customers, Birst offers some interesting new twists. By leveraging its semantic layer, search capabilities, and some pseudo-artificial intelligence, Birst Visualizer aims to bring the world of ad hoc visual analysis to non-analysts.

For example, a product manager could type "sales for Mass..." and Birst Visualizer automatically displays a line chart with the past four quarters of sales for Massachusetts. From there, users can filter and fine tune the chart with a few clicks of the mouse to detect a trend or follow a hunch. This search-driven interface, which makes educated guesses about what data users want and the form in which they want to see it, takes ease of use to the next level.

Birst Visualizer is easy to use because it rides seamlessly on top of Birst's cloud-based data warehousing platform and semantic layer. In other words, by the time Birst customers install Visualizer, they've already done the hard work of hammering corporate data into shape so users can query and analyze that data effortlessly. This is not the case with stand-alone visual discovery tools which have lightweight data integration tools but often rely on existing data warehouses or data marts to normalize and harmonize data.

Like its BI brethren, Birst bundles Visualizer for free with both editions of its software, Birst Discovery and Birst Enterprise. This means it's free for existing Birst customers, not so free for non-customers. Regardless, Birst Visualizer, which ships in February 2014, looks like the perfect analytical tool for the new breed of data-driven executive who wants to manipulate and analyze data directly instead of relying on analytical specialists.

For more information, go to this link at Birst's Web site. Also, see my 2013 report on visual discovery at www.b-eye-network.com, which compares visual discovery products from more than a dozen vendors.


Posted January 21, 2014 12:04 PM
Permalink | No Comments |
PREV 1 2