Step 4 of 10 Steps towards the Agile Enterprise

Series Recap

1. Step 1: Process Automation - Everyone Starts Here
2. Step 2: The Well defined Process
3. Step 3: The Managed Process

Answer This : 80% of your process & workflow is now automated, but how much of the business decisioning is really automated?

Lets take the Loan Origination Process in Mortgage as an example. The steps in an L.O.S workflow could be:

1. Basic Applicant Details collection
2. Pre-qualification
3. Applicant Data Collection - detailed
4. Eligibility & Pricing
5. Underwriting
6. Loan Disbursement

Most L.O.S automate the basic sequence successfully. But, the difficult part is to automate the steps that involve decision making that are driven by complex rules of the business.

For example, consider these steps, Pre-qualification, eligibility& pricing, and underwriting. These are steps in the loan process where the business will have to take decisions that are driven by company policy.

These business decision rules are very complex, varied and normally their execution is not automated. Some of these policies change at a very rapid pace. The process steps never change at the same rate that these business policies do.

The business process will account for these steps through a human interaction step. For example, underwriters will decide on the best offer the company can give for a given loan applicant, and feed result back into the system.

The automation levels for these steps are normally very low, more like 20-30% rather than 70-80% for the whole process. Most of the decisioning is still manual driven by business representatives with their rules captured in word documents or Excel spreadsheets.

The important realization is that Business Process automation does not necessarily mean that the business decisions are also automated.

Many enterprises adopt traditional IT techniques in an attempt to automate business decision making steps in their process. The common techniques include

1. Rules captured in code
2. Rules captured in stored procedures in a database
3. Rules captured as javascript in web forms

There is no problem with these techniques, but there are pain points that crop up. Some of these pain points are

1. The IT system becomes one big black box. The Business cannot look into the system and see how business decisions are being taken
2. Many business policies are very volatile. Changing them as fast as the business wants normally becomes a real pain. You never know what impact a change in your code base can have
3. Communication problems crop up, leading to delays and errors, one of the main reasons why the enterprise wanted to start automating in the first place. Intermediaries & middlemen are required to communicate the business requirements to the IT
4. Process automation has led to great improvements in productivities, but now the complexities of decision making slowly become a bottleneck
5. And finally, but not the least of problems, over a period of time, the rules get scattered all over the place, across, excel sheets, across code modules. Process and decision improvements become difficult as a big picture view becomes harder and harder to acquire.

So, what should the next logical step for the enterprise be? The process itself is well-defined, but the business policies are not.

The next logical step for the Enterprise then should be to define, document & capture their business policies that drive business decisions. That will be the topic of my next post