I spoke at the IBM Process Transformation Summit today on Transforming Business Operations One Decision At A Time. I began with some examples of operational excellence, showing how four pillars really matter – data, metrics, processes and decisions. Of these, it is decisions and processes that offer opportunity to transform business operations.
Organizations that transform themselves in this way:
- Focused on a decision
- A decision about a single interaction, a single customer, a single transaction
- Automated that decision
- Leveraged this decision automation to drive process transformation
Many organizations don’t automate decisions because decisions:
- Decisions are rich in business meaning, making them hard to code
- Decisions need business and IT collaboration and business people don’t read code
- Decisions are constantly changing but code is hard to change
To automate decisions, organizations need to adopt a Business Rules Management System to address these issues and deliver the transparency, collaboration and safe agility they need. Business rules are the baseline decision automation technology but being a decision automation hero is going to take more – technologies that can answer more complex questions are required.
Analytic and Cognitive capabilities can be added on top of a business rules platform to expand the range of decisions being automate, improve the quality and precision of decisions and drive more value.
These technologies can be added inside decision services, allowing ever increasing sophistication of decision-making without process change.When these technologies are applied to automate decisions, processes can be transformed in four ways:
- Cheaper by eliminating manual steps, waste
- Sooner by driving straight through processing
- Better by targeting precisely and uniquely
- More flexibly by providing decision agility
Recommendations?
- Treat decisions and processes as peers
- Adopt a Business Rules Management System to automate decisions
- Add Analytics and Cognitive to improve decisions