Are You Ready for Business Intelligence 2.0?
The Answer is in Data Management

A new buzz in business intelligence (BI) is business intelligence 2.0. While traditional business intelligence and data warehousing are concerned with analyzing the past, BI 2.0 concentrates on the future. Put simply, it refers to drawing inferences from historical data, applying the resulting insights to events as they happen and then managing future events through predictive analysis.

Putting this vision into practice, Accenture worked with Metro St. Louis to deploy predictive monitoring of their buses. We installed sensors to electronically monitor engine and transmission operating data so that mechanics could fix buses before there was a breakdown. Metro St. Louis realized significant savings by extending the lives of buses, optimizing maintenance schedules and identifying fuel inefficiencies.

Just a few years ago, it took weeks or even months to detect an unusual process, analyze the event, formulate and take the required actions. BI 2.0 provides these capabilities in real-time. BI 2.0 brings a burst of radical thinking and a fair number of promises that, when realized, will make a real difference to bottom lines and help companies move towards high performance levels. But what is really needed to embrace BI 2.0? Smart CIOs are examining their data management.

To receive more Research & Insights, sign up for My Outlook, your single e-mail source for all of Accenture's latest ideas and innovation, personalized specifically to your business interests and the industry issues you face.

Accenture's research has shown that 92 percent of CIOs widely include structured data in their information strategies and almost 60 percent see BI as a core component for competitive differentiation. These findings come as no surprise but disturbingly, traditional business intelligence is too often used in an undifferentiated way. Aggregated data from the past is often viewed outside of its context and compared with static key performance indicators. Knowledge workers receive standardized reports and then take time to interpret the data and make decisions.

Business intelligence 2.0 focuses on business events and how business processes and business users respond to them. For example, unusually high returns of a best-selling product would lead to the examination of many factors. Is that particular batch of product faulty, is there a pattern to the consumers who are returning this product, is there a problem with the packaging, or the sales staff, or even evidence of fraud? In this simplistic example, an "event" called "product return" triggers a series of responses that require access to information and should trigger intelligent decision making, based on a variety of conditions.

The vast majority of applications and processes have limited ability to absorb changing business needs, are not explicitly defined and do not have comprehensive metadata management processes. If we are to make BI 2.0 a reality, we first need to look at the assumptions and promises of BI 2.0 from a data management perspective.

Business intelligence 2.0 has five key characteristics, and each affects the type of data management needed.

BI 2.0 is event-driven and real-time. With traditional BI solutions, a knowledge worker is needed to interpret reports and make decisions, and so there is typically a gap between knowing something and doing something about it. BI 2.0 employs greater automation to meet its demands of speed and volume. When a business event happens, data is generated and analyzed at the same moment. Alongside this, there is also a need for real-time assessment of data quality. If we are to automate decision making, the quality assurance of the data involved should be significantly higher than for traditional business intelligence.

BI 2.0 is both individual data-centric and aggregate data-centric. We understand each incoming data point by comparing it with past aggregate data and predictions that can be drawn from past behavior. At the time of a particular event, the transactional data, business context and associated predictive models and decision-making rules could all be highly distributed and at very different levels of granularity. The technology and data governance must allow for these data resources to be marshaled in real time, irrespective of where they are kept, how they are kept, and what kind of data security and access mechanisms are in place.

BI 2.0 is forward-looking. BI 2.0 draws inferences from the current situation in order to understand likely future business impact. The resulting real-time decision making uses a series of decision trees and business rules that adapt as the scenario changes. This depends on in-memory data storage. Managing all the associated data against individual business events requires robust metadata management frameworks, especially since the nature of operational BI requires that data definitions and usage be managed and governed in a distributed manner.

BI 2.0 is process-oriented. By making BI query, reporting and metrics tools "in process" enablers, companies can root out inefficiencies relating to cycle times, productivity or operational costs. In order to gain this insight and provide automated decision-making ability, the data structure must provide for the data and the business context to be preserved together as one event leads to another in the chain of business events. Both the data attribute and the value it is carrying should be intelligent and automatically aligned to changing business contexts and strategies.

BI 2.0 is scalable. Traditional business intelligence looks at past data, whose characteristics and scope are understood. BI 2.0 looks at the ever-changing present, and so the data can be unpredictable, occurring in high volumes across multiple business events and applications. Data architecture, data definitions, metadata, usage policies and security must all be planned to handle sudden bursts of unexpected data. Systems must be able to handle large amounts of data in memory in order to quickly analyze it and react.

Business intelligence 2.0 is a more pervasive approach than traditional BI, requiring the real-time gathering of data as events happen, followed at once by analysis and predictive monitoring, which then feed back into the system, ultimately allowing automated decision making. This focus on speed and forward-looking intelligence will be increasingly compelling to organizations that can commit to getting the underlying data right.

A well thought out approach to data management and architecture must come first as a robust platform for BI 2.0 to operate. Each one of these elements needs to be part of a BI 2.0 framework or it will not reach its full potential:

• Data governance
• Data structure
• Data architecture
• Master Data Management
• Metadata Management
• Data quality
• Data security

Once data is ready for BI 2.0, the new opportunities are worth the effort. Real-time event monitoring enables organizations to attain heightened responsiveness and insight. Delivering information quickly is good, but deciding quickly is better, as it gives a significant edge over your competitors to act on the right decisions before they do.

Learn more about Business Intelligence.
Learn more about Data Management & Architecture.

To receive more Research & Insights, sign up for My Outlook, your single e-mail source for all of Accenture's latest ideas and innovation, personalized specifically to your business interests and the industry issues you face.