Data Performance Management (DPM) provides timely access to information used to drive business success. DPM consists of technologies and practices that assure business decision-makers get the information they need, when they need it, where they need it. The concept originates with Webster’s Dictionary’s definition of performance: “the fulfillment of a claim, promise, or request.” The purpose of DPM is to focus technology and practices on fulfilling the promise of data as a strategic asset within business. Assuring the performance of data means assuring that data fulfills its purpose for business managers.
The concept of DPM arises from market and customer research done by HyperRoll. Research data show that businesses now realize that current data warehouse and analytic systems no longer offer significant competitive advantages in the way they deliver information to users for making business decisions. These systems fall short because their data management architectures are based upon RDBMS systems and data warehouse practices that haven’t changed significantly since the 1980s. In contrast, DPM provides a fresh, innovative approach to the architectures and practices companies use to compete in the marketplace through the use of data.
Data Performance Management Requirements
The intense competition and fast pace of business today place pressure on information systems and their users. DPM helps them respond to this pressure, and in doing so it must satisfy the following key requirements.
It must be fast. Queries must return answers quickly so users can uncover causal relationships, trends and anomalies without waiting for the system to deliver the next iteration of data in their analytic investigation. Further, the system must be transparently integrated with other elements of the data infrastructure, such as ERP and other transaction systems, and without apparent latency in updating data.
It must be agile. Regardless of whether business data concerns inventory, customers, or employees, managing that data means accommodating changes in it – changes that often come minute-by-minute. As well as being fast, DPM systems must transparently accommodate changes in the business, in its data, and in its metadata.
It must be scalable. A DPM system must have the capacity to accommodate rapid scaling because as business and data grow, user communities suddenly increase from hundreds to thousands, data sets go from a few gigabytes to tens of gigabytes, and analytic complexities expand from a few processing operations to tens or hundreds of operations. Scalability of the DPM system is required because businesses seek to deploy analytics to large-scale operations without the prohibitive expense of super-linear scaling of hardware.
It must be available. The data must be continuously available, 24 by 7, without any apparent latency from the time the data is created until it is used for decision-making.
It must be open. The DPM system must work within an existing data management environment without requiring organizations to undergo expensive migrations or “rip-and-replace” technology upgrades. Specifically, it must work transparently with existing RDBMSs, operating systems, hardware, and security subsystems.
It must be easy to use. The DPM system cannot impede users in doing their jobs. Data managed by the DPM system should be available through any delivery means including BI reports, ad-hoc queries, dashboards, and PDAs. Because the system is easy to use, the number of personnel, the costs and the time required to manage the overall data and analytics environment should decrease when it is deployed.
Data Performance Management Technology
DPM technology is part of the data management layer, as shown below. It is not middleware, business intelligence, reporting, or dashboard technology, or an ad-hoc analysis tool. Rather, it is a peer to other data management technologies such as RDBMS, EII, ETL, and other database management systems. It addresses key technology constraints that inhibit rapid, effective delivery of data to business.
DPM differs from other data management technologies because it overcomes architectural limitations of those technologies and is complementary to them. With the addition of DPM, RDBMSs, ETL systems, BI suites and analytic applications all can be used to perform otherwise impossible analytics. Learn more about the HyperRoll Data Performance Management Suite™.