For a long time, business intelligence has taken due credit for pioneering the fight against application silos and the lack of data interoperability in corporate IT. BI analysts and planners have forced division and operational staffs to think in a larger organizational perspective by asking for data that can help consolidate divisional or cross-organizational performance views. Profitability, revenue, cost trends and project return on investment are some of the organizational views BI has helped formulate.
Of late, however, there's been a bit of a role reversal. Operational and divisional stakeholders are making an increasingly strong push for data. Now, operational systems supporting supply chain management (SCM), customer relationship management (CRM) and enterprise application integration (EAI) are forcing BI systems to react and amend their practices to accommodate organizational needs.
Enter business activity monitoring and business process management. BAM -- the real-time monitoring and analysis of processes to improve performance -- has begun to feed into, shape and demand results from BI systems. Likewise, BPM has affected and molded a number of BI systems and processes in new ways. Lets look at both of these dynamics in more detail.
Business Activity Monitoring
BAM, as an acronym, is research firms' way of differentiating an existing process to highlight a few new trends. Put succinctly, BAM is the data warehousing of key organizational performance metrics, with the twist that this data is presented in real-time through a portal or dashboard. The dashboards in this case have drill-down or actionable control functions, so users can not only monitor but also respond to and control the processes under watch. In this fashion, BAM has also picked up the moniker "operational BI."
BAM differs from data warehousing in that the latter tends to be historical and have snapshot data feeds taken on weekly, monthly, quarterly or ad hoc bases. BAM also differs from portal applications that rely on data warehouse feeds or use Web services. The real-time data backbone of BAM distinguishes it from other BI and EAI systems. In fact, as BAM matures, it picks up pointers from supply chain management and customer relationship management systems. Such systems often have portal interfaces similar to those used in BAM, but distinctly different content, plus pre- and post-processing capabilities. The key insights from SCM and CRM? Not all data has to be real-time. Sometimes data is best when it's "right-time."
There's a right-time component of data needed for a true corporate view that includes historical data, as well as operational data that's fed to the system on an asynchronous basis. If one sees the need for a lot of data orchestration here, that is precisely the point of BPM tools. BPM systems are in the business of orchestrating the delivery of data at the right time and place.
But before we look at the BPM side of the equation, it's worthwhile to consider the effects that BAM has had on BI applications already. Some of the major BI vendors such as Actuate, Business Objects, Cognos, Hyperion, Informatica and others are starting to offer more BAM-like capabilities. The same is true for portal vendors like Sybase, Plumtree, Oracle, BEA and others. And guess what -- EAI vendors such as IBM, SeeBeyond, Tibco, Vitria, WebMethods and so on are linking to dashboards to give their products more BI/BAM reach. The BAM vendors themselves are revising their offerings as well, with firms like Celequest beefing up dashboard configuration and functionality. Events and alerts, not raw data, are the driving force in a BAM system. Data-gathering, drill-downs, stop-loss signals, operator intervention and automated shutdown are some of the range of actions being built into BAM systems.
Two important trends are underway. First, BI is being operationalized against a set of business rules as one of the driving forces in BAM. And this data is being brought through dashboards to the attention of business mangers. Second, real-time data and interoperability between systems, databases, and application servers are becoming essential. In effect, BAM and BI are leading the drive to application integration.
So now let's look at how BPM contributes to this change.
Business Process Management
BPM systems capture all sorts of data about their own performance -- waiting times, average queue lengths, mean processing time. This "state of the process" data is used to monitor and manage the workflow through the system. It also provides the basis for improvements to the system. This is an area where BPM meets BAM.
Because of its self-generated process metrics, BPM systems have the same need for dashboard-like monitors, analysis functions and control functions that BI and BAM systems use, but with a broader view of dataflows. In addition, the process metrics are measured against a set of business rules to help manage process flow. A key differentiator among BPM systems is how easy they make it for managers and analysts to reconfigure the system to respond to changes observed in the workflows. This is a self-adjusting strategy that is being carried over to BAM systems.
Until recently, BPM vendors have lacked complex real-time processing capabilities. BPM vendors have partnered with BAM players, providing the needed ad hoc event-driven and scheduled data in-flows required for BAM processing. This is the second and more critical point of contact between BAM and BPM. However, as much as BPM is a partner to BAM, it is also a competitor.
In the current enterprise application field, vendors of BI and BAM are dancing a convergence tango with other technological functions including EAI, business rules management, enterprise information integration and BPM. These vendors are starting to cluster around logical enterprise-monitoring and control systems, both in vertical markets and in terms of the key horizontal feature sets. As both the BPM and BAM markets continue to grow, expect to see more consolidation across these disciplines over the next three to five years.
But even more significant is the fact that BI is leading the enterprise development and application integration parade. Not Web Services, not SOA, not ESB -- these are just the means to an end. And the end is better decision-making at all levels within an enterprise. Bottom line return on investment is what BI methods, processes and systems are delivering -- so why not lead the parade?
Jacques Surveyer is a consultant and trainer; see some of his tips and tutorials at theOpenSourcery.com