James Taylor's Decision Management

One of the potential problems with Business Process Outsourcing is that it tends to ossify processes - attempts to guarantee performance contractually at a detailed level can force a BPO to stick to the old process while some BPOs run a very standard, and very fixed, process to deliver cost efficiencies. Even when BPOs use more software and less people it can be hard for their customers to manage changes resulting in overly static processes.

Business Process Management software, used to automate a process internally, can fail the 80/20 rule - the technology can reduce costs by eliminating the 80% of tasks that took 20% of the effort. Complex tasks, requiring expensive people or lots of time, are often the hardest to automate using BPM. Lastly BPM approaches sometimes assume that any change is a process change, making change more expensive than necessary.

• If I automate a process, how do I control it?
• If I outsource a process, how do I customize it?
• When do I really need manual intervention and when can I get Straight Through Processing?
• If I run a process for many customers, internal or external, how can I make it efficient for all and yet custom for all?
• If I automate a process, how do I automate the decisions within it?
• How can I standardize a process for repeatability and avoid one size fits no-one?
• How will you show compliance?

A decent software infrastructure, and one based on SOA, is going to be critical for Business Process Outsourcers (BPOs) in the future. Business rules need to be part of this infrastructure so that the decisions in a process can be controlled by those hiring the outsourcer. The addition of decision automation to your business process automation/outsourcing efforts gives you:

• Ability to automate both data-driven and human-controlled work process tasks.
• Business rules reduce the number of required BPM tasks.
• Reduced costs/time in management of business systems by letting business owners control business components, freeing IT for higher-value activities.
• Improved competitive marketplace response through faster reaction to changing business conditions.
• Increased profitability through use of detailed customer segmentation and consistent policy enforcement.
• Ability to drive exception-handling processes directly from the rules that detect the exception.

But

• The rules must be exposed in a way that allows the customer of the BPO provider to manage the rules.
  The critical capability is that the customer can continue to control the rules, even while the BPO runs the process
• The rules must be layered
  While some of the rules will be provided by the BPO as part of its expertise and thus standard across multiple customers, those being edited by a customer must be different for each customer. To build a scalable platform this means one service taking a decision but taking a decision in a customer-specific way. Some rules engines support this approach, others do not. If you have to build a rules-based decision service for each customer, it is not going to work.
• Many decisions involve both business rules and analytics - knowledge/expertise embodied in the rules and data analysis embodied in the analytics.
  Only focusing on the value of analytics in terms of process analytics is not enough. The BPO will need to allow customers to build models from data and inject these analytics into the decisions also. Smart BPOs may realize that anonymized aggregated data gives them the power to build analytic models also and use these to differentiate the service they offer, especially to customers that do not have enough data of their own.

In the future companies will want the flexibility to automate and outsource processes differently over time and that control of the decision points in the processes, using decision automation technology like business rules, will give companies unique competitive advantages.

Prompted originally by "Creating an SOA-Enabled BPO Platform" by William Martorelli of Forrester