Posted by: Christina Torode
BI software, Business Intelligence, CIO, data management, mobile computing
The most tangible success measure of business intelligence technology is usage — and mainstream BI just isn’t there, and won’t be for some time, according to Gartner analyst Kurt Schlegel. The pronouncement came during a presentation at the 2010 Gartner Business Intelligence Summit in Las Vegas last week.
He’s predicting that this will change, however, given a boost by nine technologies that he believes will put BI usage on the same mainstream usage trajectory as that of the Internet.
Before 1993, few people used the Web, but technologies such as broadband, Web browsers and search engines changed all that. These technologies gave people ubiquitous access to information. Then Web 2.0 technologies came along, turning Web surfers into content creators, he said.
Schlegel believes emerging business intelligence technologies such as columnar databases, interactive visualization and scenario modeling, among others, will allow users to follow a similar adoption path for BI.
Here’s a rundown of the nine technologies Schlegel predicts will kick-start mainstream BI usage:
In-memory analytics: DakotaCare, a small managed health care network provider in Sioux Falls, S.D., was able to compress 140 million records with hundreds of columns of data on every claim paid since 2001 into a QlikView server. The server was on an x64 dual-core Xeon processor with 12 GB RAM.
“That is not a huge amount of memory,” he said. In-memory analytics are offered by niche players such as QlikTech International AB, as well as big BI vendors such as SAP.
Columnar databases: This lets you store data by columns, rather than rows. A columnar-based approach for data storage is better for data analysis, and, in turn BI, because it’s well-suited for complex queries of large amounts of data. Vertica Systems, Sybase Inc. and ParAccel Inc. are a few vendors in this space.
Cloud services: As BI evolves, companies will start to tap data from outside sources. He predicts that a group of SaaS providers will aggregate and offer data analytic services to fill this need in the cloud.
Interactive visualization tools: Many vendors such as Tableau Software, Tibco Software Inc. (with Spotfire) and Advizor Solutions Inc. display multidimensional data on a 2-D screen. Today, users don’t have to just look at static pie charts, but interact with them by drilling down into individual pie wedges. On top of that, users can interact with a variety of reports or heat maps and geographic maps. “These are tools that require no training — you don’t have to be brainiac number crunchers to use them,” he said.
BI integrated search: The concept: putting a search engine interface on a BI platform and being able to do ad hoc queries seems simple enough. This would really bring BI to the masses, but there aren’t many companies using this technology in production yet. Schlegel likes the idea of using the Internet as an index that spits back query results, but … “I don’t have any warm fuzzies about this technology yet. I just don’t have the [customer] references for this technology.”
Mobile BI applications: “The ubiquity of [mobile devices] makes me believe that this has got to happen.” He thinks there will be a huge explosion of analytic applications to the iPhone. For now, the most users can expect are static reports.
Data mashups: Let’s just say this is coming if Microsoft has anything to say about it. Microsoft PowerPivot for Excel comes out next month and will give users a free tool to download up to 100 million rows of data from different sources. Microsoft aside, users are going to grab hold of the ability to mash up data sources to create their own content. The best bet is to create sandboxes, or isolated areas, in which users can play and not prohibit the use of such tools, he said.
Scenario modeling: Is great for what-if scenarios: What if we moved sales to another region? What if there is an economic recession? Companies have to rely heavily on IT to go in and create alternate scenarios, but with scenario modeling, more business users can create their own what-if scenarios. Toyota is a classic example of why what-if scenario analysis is needed, he said, given its recent product quality issues.
Analytical master data management: IT typically tells the business what dimensions are being measured across a company and how they are being measured. In the future, Schlegel believes, users will be able to create their own data modeling environments and measures, submit those measures to an approval process and not have to rely on IT to make changes. Some tools that are starting to enable this capability include Oracle Hyperion Data Relationship Management and IBM Cognos Business Viewpoint.
This is a lot to take in, when many companies already have several BI tools in place and are looking to consolidate. Many are also grappling with how to get BI in the hands of everyday workers, although several of these technologies seek to address this dilemma.
Email me at firstname.lastname@example.org to let me know what technologies are on your radar.