MIAMI – One confounding aspect of information management is that, despite our attempts to dissect the topic, it always finds itself crossing defined boundaries for practical yet undefined reasons. This is not just a bane for business and IT managers told to gather their attention around a three-letter acronym (TLA); it’s anathema to the business models of conference planners who send people scurrying to different events addressing issues of business process, integration and data management -- as if all of the above didn’t need to relate to one another at some point for their clients.
BPM, SOA, BAM, BI, EPM, MDM, CEP. It’s too much, right? Not really. More often we've become used to hearing the complexities of business goals translated to reflect the technology at hand, rather than the underlying weaknesses of business models and tactics. What defines a business and the points from where it should be governed are the processes and policies that make up its daily activities. That makes TLAs mere servants to the business model, albeit useful and complex ones.
The smarter businesses I speak with don’t simply “go after” business process managment (BPM), service-oriented architecture (SOA) or master data management (MDM); they back into those philosophies with a well-understood business goal in hand. When they do that, they usually find they’re dealing with more than one component, service or TLA. Such a view offers a better way to manage scope and envision an incremental approach to a sustainable program.
Here in Miami I’d just wrapped up a visit to the user conference for Software AG, for whom I served as a judge of the company’s annual client innovation awards. Software AG, as many know, works across the process management and integration divide with software and solutions for business and IT. Lately the company is even more aligned to the market for service-oriented architecture due to its acquisition of webMethods.
Elements of BPM, SOA, MDM and business intelligence (BI) can be unified within a program as I discovered in some elegant case studies in the awards process and in live demonstrations. Here’s a brief explanation of one presentation I witnessed from an end user as I interpreted it:
A brand name online retail broker (I didn’t have time to clear disclosure issues) needed to improve order cycles across channels of self-service, call center and field sales. The legacy process had a cycle time of two to four days, a support staff of 10, operated only during business hours and had an unacceptable rate of data errors.
The goal was to increase revenues through better online inventory management and same-day fulfillment with a lower rate of data errors. Beginning in 2006, the company began to reengineer and align business processes across the three sales channels and instantiated a targeted SOA and MDM strategy. During 2007, multichannel data was transformed into master records for customer, inventory and dealer. Through various process points, an enterprise service bus (ESB) is used to feed master data back to the three sales channels. From the same process points, data was aggregated and made available to business intelligence analytics for review and action.
At this particular company, business analysts and IT staff work together to design, test, deploy and monitor business processes with an emphasis on collaboration. Over time, they say, it will reduce TCO and increase availability. The company claims that since the program was begun, many processes that required human intervention are now automated, cycle time has been reduced to one day, data quality errors have been reduced by 20 percent, and the system is available to customers 24 hours a day. Benefits in 2008 include more power extended to business intelligence users through operations and process visibility, analytics and real-time trending.
This summary is short on detail, and I’d like to follow up with this company to clear approval for more of what I heard for a case study. An initial exploration of the company’s services didn’t disprove anything I saw. But the point of this example is to show that targeted programs encompassing, in this case, BPM, SOA, MDM and BI are more understandable and digestible when they all point to a problem at hand.
