SaaS Business Intelligence
Group Options
Members
Self-service BI
The move toward self-service business intelligence, for instance, has been picking up steam and will gather further momentum in 2011, Kobielus said. Increasingly, enterprises will adopt new Web-based interactive querying and reporting tools that are designed to put more data analytics capabilities into the hands of end users, he said.
The tools will give end users the ability to quickly navigate through and visualize business data, and they will allow them to generate views and reports relevant to their job functions.
Numerous vendors, including IBM, SAP, Information Builders, Tibco Software, QlikTech, and Tableau Software, already offer such tools, and adoption will accelerate as more companies try to deliver BI capabilities to nontechnical users, business analysts, and others, he said.
Self-service BI tools "take the burden off IT and speed up the development of reports to a considerable degree," Kobielus said. They also make it easier for users to create personalized reports that reflect their needs better than the standardized reports developed using traditional BI approaches, he added.
SaaS BI
The increasing availability of BI tools provided on a software-as-a-service basis will also drive more self-service BI and enable wider adoption of BI usage in general, Kobielus said.
One example of a company that is taking advantage of SaaS BI is New York-based women's apparel maker Bernard Chaus Inc. The company is using a BI application to track the performance of its products at each of the high-end department store chains that sell them.
Every week, company executives sift through sales data from each department store to see how each of its SKUs are performing and determine which stores might be doing a better job of selling them. The data also is used to study which styles and designs are popular and are selling well.
That sort of analysis is vital said David Stiffman, chief operating and financial officer at Bernard Chaus. "Reading consumer response to certain styles helps influence what we will be designing in future," he said. "By analyzing and learning what sells we are able to make better decisions about what we will offer customers and how to encourage them to buy from our line."
But rather than host a BI application itself, Bernard Chaus has signed up with IBM business partner Sky IT Group, a firm that offers SaaS BI services. Sky IT brings in data from all of the different department stores that sell Bernard Chaus apparel and then cleanses the data and makes it available to Bernard Chaus for analysis.
The approach has helped Bernard Chaus take advantage of BI tools at a substantially lower cost than the price of doing it in-house, according to Stiffman.
A slew of big and small vendors have begun rolling out such hosted SaaS offerings, and more companies will start offloading at least some of their BI applications to such vendors, Kobielus said. In many cases, SaaS BI vendors are able to "provide BI at a lower cost so companies don't need to maintain the staff or manage the footprint," internally he said.
Hadoop
In another development in the BI market, expect enterprise interest in the open-source Hadoop framework to increase significantly next year, said David Menninger, an analyst at Ventana Research.
Vendors such as IBM, Pentaho, Cloudera and Karmasphere already offer enterprise-ready Hadoop implementations, and analysts expect more vendors and more products to join the list in 2011.
The reason for the interest is Hadoop's ability to let enterprises analyze much larger volumes of data than most traditional database systems and warehouses can comfortably manage. Much of Hadoop's growing popularity also stems from its usefulness in social media analytics and text-mining applications, Menninger said.
"Hadoop is the new black. It is gaining in popularity because it supports a wide variety of analytics and [data types] that we couldn't previously analyze either because the data was too big or the analysis was too complex," Menninger said.
"Not everybody understands it yet, but Hadoop is going to have a big impact on big data infrastructures [and social media analytics,]" he said.
Open-source BI
Kobielus also expects that other vendors of open-source tools, such as Pentaho, Infobright, Jaspersoft, Talend and LucidDB, will start rising in visibility and will soon start to offer more complete BI stacks.
Pentaho, for instance, already has one of the stronger commercial Hadoop capabilities, and many of the others offer innovative technologies that are resonating with users, he said.
For instance, Bango, a Cambridge, England-based provider of mobile analytics and billing services for large content providers, started using Infobright's columnar database technology when its older SQL Server-based database began struggling to keep up with exploding data volumes.
What makes the technology appealing is its ability to support ad hoc complex queries on large data sets, without any need for any indexing, manual tuning or IT support said Tim Moss, chief data officer at Bango. "This is huge," he said.
Infobright stores metadata about data even as the data is entering the database. "This means you don't need to index the tables; it's kind of doing that automatically," he said. "With other databases, you put the data into tables and then, depending on queries, you need to add indexes to help speed up and support these queries."
Real-time analytics
Expect enterprises to pay more attention to products such as SAP's High-Performance Analytic Appliance, or HANA, that are designed to speed up data analytics, Kobielus said.
HANA uses an in-memory computing technology that allows data to be processed in a system's RAM as opposed to reading it off I/O disks. The in-memory approach enables much faster data processing and is designed to allow companies to run far more sophisticated data analytics applications than they could with conventional relational databases.
For the moment at least, such in-memory technologies are considerably more expensive than traditional disk-based products, but expect that to change as the technology matures and more people start using it, Kobielus said, adding, "BI is becoming more real-time."
In 2011, expect to see enterprises continuing to push for "all things cache, all things memory," Kobielus said. In-memory and flash-based technologies are slightly ahead of the curve, at the moment, he said, but added: "I think that will start changing in 2011 and 2012."
Emerging new search and discovery technologies from companies like Attivio and Endeca will also start making a bigger impression in enterprise environments next year.
Such tools are designed to enable companies to implement user-configurable, search-based business intelligence applications involving large volumes of structured and unstructured data.
"Search tools are getting embedded into the BI stack" and are enabling a convergence of unstructured and structured analysis, Menninger said.
SOURCE: Infoworld