Massively parallel database drives bigger BI

Busy managers are making more complex decisions in less time and, flooded by data, leading big corporations are moving toward enterprise-wide business intelligence (BI). Those are among findings of research carried out for massively parallel database and BI company Teradata. The firm’s figures show data rates compounding, with 97% of respondents seeing serious increases, and over half saying data has doubled or tripled over the previous year, while more than 50% say decisions are more complex this year than last. “These findings tell us why companies today are moving enterprise analytics to the top of their priority lists,” says Bob Fair, Teradata’s chief marketing officer. “The overwhelming majority of respondents, more than 70%, say that poor decision-making is a serious problem for business.” He also notes that Teradata’s research indicates the power of centralised enterprise data warehouses – with 75% of those doing well in terms of decision making running with the technology, while only 19% of those struggling do. However, most firms below the large corporates still don’t understand what enterprise data warehouses are or what BI can do for them. Rusty Warner, director of product marketing for Teradata in the UK, says education remains a serious obstacle. “Business intelligence isn’t just reports and what happened last week. Companies need to understand that they can perform forward-looking functions with good data warehouse functionality – and they can do it on the fly,” says Warner. “It transcends traditional business boundaries and limits. It also means they can save money by not having departmental operational data stores. Redefining business processes should be at the top level of the chain of command.” Both were speaking after the launch of Teradata’s Warehouse 8.0, which boosts performance beyond traditional decision-support, with enhanced integration into enterprise systems, new real-time data management facilities and access to vast amounts of business, engineering and manufacturing data. Warner says it means that far more people in large manufacturing organisations can run more complex operational and strategic business analysis, taking into account more data and more data types in real time. It also includes facilities to integrate history and trend data with current information to initiate event-based real-time actions. And it includes new facilities. Warner gives an example of big automotive or aerospace OEMs wanting to run recursive queries on complex assemblies. “They would normally have to build temporary BoMs which alone might take eight days. With the new system, they can pull that data out on the fly: it’s a database function that enables them to look straight into very complex large scale assemblies.” Operational data too can be better integrated into the data warehouse, with triggers calling and running procedures to improve, for example, customer service. “It’s good for companies trying to improve order fulfilment, or wanting to update total forecast on the fly for their supply chains,” says Warner. He makes the point that where data is normally held on several systems: it can now be in one place, meaning that everything from business impact analysis to front-line operations can be supported better and more quickly. Teradata has APIs with SAP ERP at several levels – for example, with BIW and also via Netweaver. There are interfaces to other systems, but Warner describes, for example, PeopleSoft as “ongoing work”, and Oracle as “a little more complicated for us.” Teradata works with most of the BI software vendors, including SAS Institute, Business Objects (and thus Crystal), Cognos, Microstrategy and Hyperion.