Banks Need More Than Workflow

by FICO

(Posted by guest blogger, James Taylor)

I was pointed to a new site this week - Gonzo Banker - and in particular to this article on Workflow: The Killer App. While I enjoyed the article, I do think that the example used shows the weakness of a workflow-only solution. The article begins with some nice definitions of the two BPMs:

The implementation of workflow technologies is the foundation of business process management. Measuring the effect of implemented workflow is a key aspect of business performance management.

I would add that adding intelligence to workflow is the purpose of enterprise decision management and just as important as the two BPMs. This is clear even from the example used to illustrate the power of bpm.

Example: Tom, a CSR, requests to rebate an NSF charge that requires approval from his branch manager, Susan. The required data is the account number, the date of the fee, the amount to be rebated, and the reason for the rebate (required for performance management, not process execution).

In the old days, Tom might draft an email, send it to Susan and ask her to approve and forward to the Deposit Services email group. In the really old days, Tom printed a form and had Susan sign it and route/fax it. Either way, the information on the form could be incomplete. By using a workflow engine, Tom cannot submit the request until all data is entered.

I would look at this differently because there is a decision in this sequence - should a rebate of the NSF charge be given. The assumption in a workflow-based solution is that this must be made by a person other than the person gathering the data. But the person to whom it is being referred is not going to collect any more data which means that all the information required to make the decision is gathered at the first point of contact. Instead of thinking about workflow, think about the decision:

Use business rules to make sure that the necessary data has been collected.
Automated a decision to rebate or not based on the bank's rules, the relevant regulations and (possibly) predictive models for fraud risk.
This decision service can use data without exposing it to the CSR and so privacy rules are easily enforced.
Immediately give the CSR the answer and reasons for it.
Collect the rules and model execution data so that the supervisor can review them later if she wants.

In this case there is no need for workflow - why move work around when you can decide and then act? Taking an EDM approach in this case means that the supervisor can spend her time on more productive activities than rubber-stamping approvals and that the CSR can give a customer immediate feedback - no open ticket, no customer waiting, no workflow!

Banking, being highly regulated and very risk-centric is full of these kinds of decisions - ones that used to be referred to a manager but that could, in fact, be automated with the right approach - EDM. Here are some other posts on banking and financial services that will give you some more examples of what I mean: