The term ‘BPM’ has been adopted in the marketing communications of just about every IT vendor and management consultant, as what comes after the dot-com fiasco. It seems everyone selling IT products or management consulting services has put BPM lipstick on their products and services. Even the IT and financial analysts are having a field day defining BPM to mean whatever they want it to mean.
Although confusion abounds, a handful of useful working definitions can bring about some badly needed clarity. Figure 1 provides a backdrop for our discussion.
The business process is the end-to-end coordinated set of collaborative and transactional work activities carried out by both automated systems and people to produce a desired result or achieve a goal. While this definition is simple enough, the coordination of complex sets of activities carried out by independent work participants (humans and machines) is by no means simple. Business processes are essentially human phenomena, but in today’s automated and wired world, automation and technology assistance to amplify human work are indispensable. This technology assistance involves automation of routine transactions and dynamic human collaborations, as suggested in the figure.
In his classic work on value chain analysis, business strategy expert and Harvard professor Michael Porter classified work activities into two types. Primary activities are those that directly touch the customer, while support activities are those that are primarily administrative, keeping the rent paid and the lights turned on. Using Porter’s framework, it’s the primary, end-to-end customer-touching processes that are paramount to gaining and sustaining competitive advantage. It is these primary business processes that deliver value to customers, and they are a company’s value-delivery system—all the rest, the support activities, are costs.
While support activities require management attention as means to make a company more efficient, primary activities require laser-focus to make a company effective in the marketplace.
Business Process Automation (BPA)
BPA simply means automating a business process. Since the advent of the first commercial computer in 1951, companies have embraced technology to streamline and speed up their support activities for greater operational efficiency. As far as primary activities, today a number of technology families have emerged for this purpose, including workflow systems and, more recently, Internet technologies such as application servers. While just about any technology can be used to automate business processes, the real issue companies face is how to optimize and manage those processes after they have been semi- or fully automated.
To date, business process automation (BPA) requires delving in technology plumbing and diddling with complex computer programs for each instance of business process change. That means that technologists, not business people, are required to create or change an automated business process. Companies are so bogged down in expensive and tedious efforts to rewire technology for each process change that, often, needed business change is simply foregone, as it isn’t practical or feasible. Thus, business process automation falls short in meeting the needs of time-based competitors and is unable to take companies much beyond where they are today, stuck with rigid business automation systems. Even with a veneer of business process modeling tools made available to business people, technical staff are still required for business process change in most BPA scenarios. But, we’ll say it again, the ultimate challenge of business process management is not just the automation of business processes, it’s their management thereafter, for change is a certainty. Thus, companies want to go beyond business process automation and on to business process management without having to run through the technology gauntlet.
Business Process Management (BPM)
BPM is a business discipline or function that uses business practices, techniques and methods to create and improve business processes. From this general definition, just about any process improvement discipline or activity, including reengineering, TQM or Six Sigma quality methods, outsourcing and lean manufacturing, can be considered as BPM. Thus, from an extremely general perspective, BPM has no distinguishing definition at all; it’s just about anything that contributes to process improvement—it can mean whatever you want it to mean.
On the other hand, the term BPM has been propelled onto the front pages of the business and technology literature for far more specific reasons. Whether manual or automated, companies have learned that the piecemeal process improvement methods and techniques they have scattered throughout their organizations don’t produce breakout results. So, if your manufacturing division uses Six Sigma, and your marketing department uses voice-of-the-customer techniques, without talking to one another, that doesn’t mean you have a process-managed enterprise, the kind of enterprise that produces outstanding and sustained business results.
BPM in its contemporary context is a holistic vs. piecemeal approach to the use of appropriate process-related business disciplines that are used to drive business performance improvements, not just across the departments in a single company, but also across multi-company value delivery systems. This approach has only now become practical as a result of the new category of BPM software systems.
BPM systems are technologies designed for the complete management, not just the automation, of business processes, from creating innovative new business processes, to their redesign and improvement over time. BPA, the one-off automation of a business process, is only a part of a complete BPM system. It’s the “M” in BPM that really counts, for change is not a one-time affair.
BPM systems provide computer assistance for supporting all process work such as Six Sigma initiatives; mergers and acquisitions; overriding processes embedded in ERP systems; implementing industry-specific collaboration protocols such as Rosettanet Partner Interface Processes for the IT industry supply chain; or complying with regulations such as the Sarbanes-Oxley Act.
Automated support systems are absolutely necessary to deal with today’s complex business structures and to get beyond the piecemeal approaches to process improvement. Without this automated support, it would take armies of people to take on the real tasks of real-world process work.
One approach to building BPM systems is to federate preexisting technologies (including workflow, application integration brokers and business rules engines) and add them to an already complex technology stack. Today that technology stack is organized under the banner of application servers, software that’s part of a three-tier distributed application: the user interface, the business logic, and the back-end databases. Applications servers are often labeled as Internet servers, for their primary use is to connect preexisting systems such as ERP and newer Internet-oriented applications. BPM servers add yet another layer to this already complex technology stack.
The situation is similar to that of the emergence of database systems a couple of decades ago. What are known as hierarchical and networked database systems used pointers and links to keep data elements tied together logically although their physical storage was in quite a different form. They worked, up to a point. When the links were broken, business information systems failed and had to be recovered by reloading the databases. That was certainly no way to run a company’s business-critical information systems.
So, along came an innovation. Instead of links and pointers, a researcher at IBM Watson Research Center, Ted Codd, used the mathematics of relational algebra as the foundation for a new kind of database system that obviated the need for complex links and pointers. The relational database management system (DBMS) was so solid and reliable that a whole new breed of applications called enterprise resource planning (ERP) systems emerged that are now the foundation for automation in large and mid-size businesses across the planet. The same level of innovation is needed for business process management if we are to get beyond the already complex technology stack and the additional complexity BPM servers introduce. That innovation is the business process management system (BPMS).
The BPMS Innovation
As before with Ted Codd’s innovative approach to data management, Cambridge University’s Turing Award winner (the Nobel Prize equivalent for computer science), Robin Milner, took a fresh approach to process management. Like Codd, Milner turned to the mathematics (pi calculus) to find the needed underpinnings for representing and manipulating processes. A DBMS is essentially about storing, processing and managing data. Similarly, a BPMS is fundamentally about storing, processing and managing business processes.
The BPMS is a radical simplification for process work, for it does not add yet another layer to the already complex technology stack—it is simply a user of the existing technology stack. Its core capabilities are built from the ground up around business processes, not technology processes. Thus, the tools built atop the BPMS can be designed for business analysts, not just technologists. With BPM servers, the business process is shredded and stored among the various federated technologies. With the BPMS, the business process remains whole, and overcomes the fragility inherent in BPM servers that is similar to the fragility that broke the hierarchical and networked database models. BPMI.org’s Howard Smith elaborates, “Business processes are long lived, they exist over time, and thus, have to be persistent. You have to be able to switch off the system and switch it back on again and it’s all there. And, processes don’t just execute, you need to manage them as they evolve, one way or another, and to differing extents. A BPMS provides a clean model for the stored processes, so that tools can be built above for managing them.”
The BPMS is of the process, for the process and by the process; not of, for and by the technology stack. In the early phases of BPM work, companies can get by with BPM servers, and their technical people will be comfortable with the familiarity of that which they already know. Business people will continue to be baffled by the technology stack. But once companies get beyond their initial BPM projects and their process repositories grow exponentially, the complexity of linking preexisting technologies with a BPM server will cause them to hit walls similar to the walls hit by early link-centered database approaches. As companies move beyond the initial tinkering phase of BPM, the need for a native business process technology foundation, the BPMS, will become all too clear.
Extending the technology stack with BPM layers is similar to trying to build a skyscraper by stacking a thousand doghouses one atop the other. Skyscrapers need an architecture all their own; business processes need an architecture all their own. The BPMS provides the process-oriented architecture needed to build a process-managed, real-time enterprise.
The leaders of a process-managed enterprise recognize that it’s the end-to-end, enterprise-wide business processes that make or break the company. They have a holistic, not piecemeal, definition of enterprise BPM. They also recognize that end-to-end process management is complex and requires not just great management theories, but also requires commensurate support from automation tools.
The process-managed enterprise requires BPM systems to realize the goals they set for holistic process management work. Companies that adopt this perspective often interchange the terms BPM and BPM systems, for the two go hand in glove if management intent is to be translated into execution.
Leaders of a process-managed enterprise understand work-oriented business architecture and design work structures around the end-to-end business processes. They know they need to bring the right tools to the table in order to realize their vision, for business is no longer conducted manually or through time-delayed, batch-oriented information systems. Today’s executives want the tools and infrastructure needed to become time-based competitors, competing on cycle time, product design time, lead time, time to market, response time, just-in-time inventories and up time. They know that to place time on their side, they need tools that can allow them to create and manage flexible and fine-tuned business processes that can keep up with and even anticipate ever-changing customer demands. Such processes must cross company boundaries and coordinate the multi-company work of the entire value delivery system.
Howard Smith explains the challenge of coordination of work. “In the largely vertically integrated companies of the past, business processes were once thought of as those that could be rigorously scheduled around well-defined roles in carefully designed workflows—routing of work from one role to the next, work that was waiting for a telephone call from a customer, or work that had to be processed at a specific time (“I will expect your call at 10 o’clock”), or work that had to be transferred to a different person because the person who did the first part of the processing got sick or quit before the task was complete. But in today’s global, horizontally integrated companies, coordinating business processes is neither as simple nor as linear as portrayed in the tidy world of traditional task management or the overly simplistic process re-design around case workers, typical of the reengineering efforts of the past decade. Today, work management is about coordination, collaboration, negotiation and commitment. Business is constantly changing, messy, unordered and chaotic, and both manual and automated work activities have to progress in parallel. Work is conducted, and coordinated, at all levels, through choreography and orchestration.”
Because the multiple companies involved in a single value chain operate on their own clocks, coordination and synchronization can only be accomplished with a robust information chain that provides the actionable information each participant needs to optimize the overall flow of work. The ability of systems-of-process to publish or subscribe to real-time transactions, occurring within active business processes, enables organizations to respond to events, such as fraudulent activity in banking or insurance, in real time. This capability also enhances business decision making for time-based activities such as situation-appropriate cross-sell opportunities, based on real-time credit card activity and customer interactions in the contact center.
Automated techniques such as creating a single, complete representation of a customer across all channels and lines-of-business, complex account opening, change of address across multiple lines-of-business, or managing lost or stolen credit cards, are example capabilities of real-time systems-of-process.
According to consumer relationship management expert and CEO of Chordiant, Stephen Kelly, “This capability gives organizations the flexibility to change business processes in a matter of hours, rather than months. The ability to liberate business processes, business policies and legacy data provides a unique opportunity to deliver business value and to realize substantial reductions in operational costs, increased retention rates and selling opportunities.”
Howard Smith elaborates, “BPM takes change off the critical path of innovation by creating a new contract between business and IT. Instead of asking IT to implement specific processes, the business contracts with IT to provide a BPM service, extending process design, deployment, execution, monitoring and optimization tools directly to business users. As companies once digitized critical business data using database management systems, companies are now digitizing critical processes using business process management systems. This is the New IT that GE has dubbed ‘digitization, a revolution representing the greatest growth opportunity the company has ever seen.’ The digitized processes and their management systems become the new platform upon which the real-time enterprise is built, as ERP systems were built atop enterprise data models and databases of the last decade. In this environment, companies are linking familiar process mapping tools to powerful process execution systems to create a new form of organizational knowledge—explicit, executable, actionable and adaptable work design. They are encoding best practices gathered from wherever they can find them, inside or outside the company, and mixing them with unique business processes, their distinguishing ways of doing business. The objective is organizational learning translated directly to operational innovation.” In fact, the learning process is, in and of itself, the ultimate business process.
Companies want to shift their efforts from further automating individual tasks and move on to managing end-to-end business processes—the very essence of the value proposition of the real-time enterprise.
This article is an extract from The Real-Time Enterprise by Peter Fingar and Joe Bellini.