Shari would like to thank Soumendra Mohanty, senior manager with Accenture Information Management Services, for his contribution to this month’s column.


Any change in thought process, concept or technology comes with an inherent notion of helping businesses perform better. In this context, business intelligence (BI) 2.0 signifies a radical change to the traditional BI approaches. It is used to describe intelligent processes that respond in real time, based on analysis of both data and events. Put simply, BI 2.0 refers to drawing inferences from historical data, applying the resulting insights to events as they happen, and then controlling future events through predictive analysis.


So should companies be getting excited about a version of BI that appears to promise much for the bottom line and help companies achieve high performance? Smart CIOs have been asking one important question: what about the data itself? They can see the promise that BI 2.0 has but also note a potential issue with the levels of data management needed to make it a success.


BI 2.0 is a more pervasive approach than traditional BI, requiring the real-time gathering of data as events happen, and then building in analytic capabilities to permit predictive monitoring, which then feeds back into the system, ultimately allowing automated decision-making. But to get to the starting point, teams from the CIO’s organization will attest to the fact that event and data go hand in hand. So to realize BI 2.0’s full potential, we first need to look at data management to ensure the concept can be realized - DM 2.0.


Fundamentally, business interaction is about business events that occur and how business processes and business users respond to such events. For example, how intelligently a business responds to the unusually high returns of a particular product which is also a best-selling product depends on many factors. Is that particular batch of product faulty, is there a pattern/correlation within 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. Event monitoring systems have proven their worth in tracking these business events. However there is also a much greater need, not only for monitoring and displaying information, but also to anticipate the need for decision-making based on the business context and the business strategies drawn from historical data.


Traditional data warehousing and BI have always been employed to analyze past performances of a business. BI 2.0 adds a different dimension to this by proposing information agility as the next leap so that as business events occur, they can be analyzed and actions can be taken at the very moment of the business event. The data and data management processes enable a robust platform for BI 2.0 to operate. Data management provides key services such as data governance, data architecture, master data and reference data, metadata, data quality and data security. Each one of these data management services need to be part of BI 2.0 framework or it will fail at the first hurdle.


Managing data and business context together in real time and ensuring the data structures are aligned to handle different levels of granularity and completeness of data is the key. Additionally, having robust data quality processes and interfaces to deep dive into data anomalies in real time is paramount because any wrong inferences might lead to erroneous decision-making, negating the usefulness of BI 2.0. Equally important is to look at some of the other areas that would heavily influence the DM space - these are process-aware BI and search-based BI.


These concepts challenge the traditional data management approaches in a multitude of ways. For example, in order for BI 2.0 to become pervasive and near real time, it needs dynamic context management of the data attribute as well as possible values it can carry based on several business event scenarios. In order to provide real-time alerts and automated intelligent decision-making abilities at the time of the business event, predictive models, business rules and decision-making trees need to be part of the business event management framework as in-process executables rather than out-process reference mechanisms. A new level of robustness and sophistication is required for the metadata management approaches to have the data, the domain values, the business context, the business rules, etc. all infused together and accessible to the entire enterprise application landscape.


Alongside this, there is also a need for real-time assessment of data quality. Transactional systems have been exposed to various levels of data quality threats, and it is extremely challenging to ensure that there is completeness and accuracy of data as it gets generated. In contrast, if we have to significantly reduce the latency in decision-making, the quality assurance of data involved in the decision-making should also be significantly higher than for traditional BI. It is clear that BI 2.0 needs a robust real-time data quality framework in order to alleviate the uncertainties around quality of data while also providing instantaneous options to correct data points yet still keeping the business context intact.


Often spoken about and misunderstood, data management services must be at the heart of any program that seeks to reap the real benefits of BI 2.0. Data management should not be treated as an afterthought; rather, it should be a key aspect of every system and application involved. Only then can companies expect to steal the lead on rivals as they operate a system that uses intelligent processes to operate real-time responses to business events using analysis and predictive monitoring. Would any CIO turn down that sort of promise?

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access