Defining cross-functional processes, such as order-to-cash or procure-to-pay, is the subject of a lot of discussion today. These processes are executed by multiple applications that are, at best, partially integrated. What happens when exceptions are flagged while responding to, for example, an order that exceeds a customer’s credit limit? An alert is raised to which an information worker must
respond. Exceptions tend to occur over and over again, so they must be managed by information workers using tools such as phone, fax, email, instant messaging — in particular, Microsoft Office productivity tools. However, productivity tools provide no assurance of consistency in how these tasks are handled, even though this type of exception is likely to recur on a regular basis.
Companies rely on information workers to be the glue that fills in the gaps in end-to-end business processes. These “gaps” represent the repeatable, operational decisions that information workers are alerted to and to which they must respond. Since these short exception-handling processes repeat over and over, they must be handled consistently to achieve both greater efficiency and better compliance. Such decision-oriented applications for information workers represent the next major frontier for business-process automation.
“The Move to Intelligent Applications,” below, shows the move to intelligent process automation (IPA) as the latest step in the evolution of enterprise applications. It is the next logical step in a journey that began with the automation of horizontal processes from core financials to Material Requirements Planning (MRP), then went to application suites and Web applications.
|The Move to Intelligent Applications
IPA software automates repeatable, operational decisions within those business-process sets where analytics drive the workflow. Software solutions that enable IPA are beginning to emerge in both packaged and custom-built forms. These solutions exhibit a variety of characteristics such as the following:
- Continuous in-process, rather than after-the-fact, data integration
- Run-time event-driven capabilities, including Business Activity Monitoring (BAM)
- The ability to optimize straight-through business-process workflow based on advanced analytics or statistics and data mining
- Highly automated exception-handling, when needed, triggering further specific process flows where analytics aid in or automate decision-making while other workflows continue
IPA addresses three major business imperatives: compliance, performance management, and the streamlining of multi-organizational connectivity:
1] Compliance is a top priority for many organizations in light of increased regulatory controls across more industries than ever before. For example, compliance decisions could include when to recognize revenue or to whom to extend credit. IPA focuses on driving consistency in the way repeatable, operational decisions are made — within and across organizational boundaries — from your customers to your suppliers.
2] Performance management: Dashboards are now widely used to measure and manage business performance with key performance indicators (KPIs). But a dashboard that simply reports on what has happened is of limited use. A dashboard that shows the current state of a business process (i.e., business intelligence, or BI, about a business-process set) and how it deviates from established norms is far more valuable.
Providing a process context for business-process management (BPM) will be a major stimulus for developing and deploying IPA. Including the performance of your entire supply chain, as well as your own organization, is key.
3] Efficient multi-organizational interconnection. Both of the imperatives above are most often executed in one legal entity. The demand is to get compliance, performance management, and other business-process sets working across organizations the same way they work within organizations.
IPA software makes it easier to automate by simplifying the way you link process sets together among the entirely different code sets, data sets, and infrastructure that are involved across different legal entities.
How Does IPA Work?
Incorporating analytics that select among decision alternatives within a process can yield measurable improvement, as shown in “Business-Process Flow: IPA,” below. This example illustrates a business-process flow in an IPA-enabled system.
|Business-Process Flow: IPA
For example, let’s say a shoe manufacturer receives a big order from one of its retail customers. Unfortunately, the order exceeds the customer’s credit limit. Should the manufacturer increase the credit limit or turn down the order?
Orders above the credit limit are not one-time events; they recur regularly. Hence, IPA, which deals with repeatable operational decisions, is appropriate in this case.
“Business-Process Flow: IPA,” above , shows the straight-through processing defined for normal order-management transactions. Let’s assume that an exception has been defined for the above-credit-limit transaction. The transaction flow, Workflow A, is monitored and when a specified event condition is detected, an exception is declared. Control is then turned over to another process flow, Workflow B, which has been developed to handle this exception. If no decision is needed, the exception invokes the prescribed rule; if a decision is required, one of three possible scenarios can occur:
1] The exception transaction can be turned over to a human credit manager to make the decision, probably with some specific analysis of the various elements in the credit history of the customer.
2] When multiple stakeholders’ inputs or decisions are required, either a structured (workflow) or unstructured (collaboration) resolution process involving several human decision-makers is triggered (still in the process context, and ideally still with the guidance of analytics).
3] In a fully automated IPA environment, analysis leads to a precoded decision without the need for any human intervention.
In both cases 2 and 3, other transactions continue to flow straight through the workflow and, when the result of the exception process is known, it also continues. A record is kept not only of the transaction, but also of the decision that is made and, later, the outcome of that decision. The decision and its outcome are analyzed and used to adjust the rules and decision criteria appropriately, incorporating learning and ongoing improvement.
This IPA example incorporates BPM to orchestrate the process flows and BI to provide analysis at the point of the decision (as well as all the underlying data and logic integration and management software). Such a converged solution extends the capabilities and overcomes the limitations of traditional BI and traditional BPM.
The SAP NetWeaver Business-Process Platform
As a platform for IPA, SAP NetWeaver’s runtime components form a strong base; they include SAP NetWeaver Application Server (SAP NetWeaver AS), SAP NetWeaver Exchange Infrastructure (SAP NetWeaver XI) integration broker, and the business-process manager within the SAP NetWeaver application platform and process integration layers.
Version 2004s of SAP NetWeaver contains a unified runtime for executing guided procedures, collaboration tasks (Outlook-like), Business-Process Execution Language (BPEL)-based straight-through processing, and hard-core workflow modeling. This technology is currently deployed to coordinate and automate simple and complex intra-corporate and inter-corporate processes.
Other key business-process platform components include:
- SAP NetWeaver Portal, which serves as a key interface for process design and execution
- Task management’s universal Worklist, which is an alert inbox that also provides the context for quick issue resolution
- Alert framework for SAP NetWeaver, which is a central element of the SAP NetWeaver architecture, providing the essential capabilities to invoke alerts and trigger follow-up processing, and offering other essentials such as escalation paths, based on response times
At Sapphire 2006 in Orlando, SAP and Microsoft unveiled Duet, software for information workers that enables access to SAP ERP from within the Microsoft Office environment, initially from Outlook. This product is only the beginning of new kinds of development. I expect to see structured applications built out of Duet and enterprise services that will automate decision-oriented processes. There will also be guided procedures to walk through the necessary steps for successful process execution across information workers and workgroups.
IPA automates the movement, analysis, approval, and action steps required to support daily, repeatable, operational decisions whether they are exception-based, straight-through, or both. However, for that capability to be meaningful and cost-effective, the user needs to consider multiple business processes — business-process sets — rather than simply one-off business processes.
This approach is a challenge in today’s IT environment because companies are looking for quick returns on investment (ROIs) — often in less than a year. However, thinking in terms of business-process sets can reach across an individual’s entire organization and out to the customer and supplier ecosystems. When this happens, it typically means cultural changes for your organization, as well as training and other related needs that simply cannot be implemented within a single budget year.
The solution is to think in terms of a highly flexible IPA infrastructure where everything is considered to be a service, in the computer science sense of the word. Such an approach allows you to switch out the old and upgrade capabilities much more quickly.
Similarly, where there is no competitive advantage to automating a business-process set, you should use one of the many industry-specific templates that are emerging both from IT suppliers and industry groups (for example, ACCORD for insurance). As a result, users can effectively deliver IPA that crosses application and organizational boundaries and does BI on-the-fly to enhance your customer service, improve staffing and workflow, and better support partners and suppliers.
|Henry Morris, Ph.D., is group VP and general manager of integration, development, and application strategies for IDC.