The Strange Case of the People vs. Business Decision Management

Essays I have authored in the past generally focused on various methodologies and technologies revolving around business decision management – business process, business rules and analytics. I’ve always been quite interested in the past, present and likely future of these capabilities and their application in various real world domains.

For the last several years I have been wholly involved with applied decision management in the financial services arena. While there has been an almost unfathomable amount of turmoil in these businesses over the past several years, I was still very surprised to recently hear a loud cry from the mortgage industry that “technology got us into the trouble that we are in today!” 

Why blame the technology?   Automated underwriting - the same technology that drastically reduced origination cycle times, drove new efficiencies, reduced costs, mitigated risks, and brought quality and consistency to decision-making.  This technology set is largely based on decision management platforms.

Shockingly, a recently published article in National Mortgage News discussed banks and credit unions “turning off automated loan decision tools in favor of manual process”  because they were no longer “trusting technology.” The same article suggested there has been nearly a 50% reduction in loans that came through automated channels.

Naturally my first reaction was disbelief. How could these approaches I consistently advocate now seemingly be the root of all evil?

Some of the reasons given for the escalating abandonment of automated underwriting and other decision-making technology appear (at first) to have at least some merit.  They include: 

• “The systems are just too easy to game; loan officers quickly figure out what data to enter to get an approval.”
• “Automated underwriting systems ignore the subjective factors that every good lender once used to assess the reliability of a borrower.”
• “The current automated decision systems don’t meet all of my needs.”

However, even a cursory examination of these issues quickly suggests the technology is really not the culprit here. More importantly, I think it tells us where we need to head going forward – in this domain and many others.

The first issue raised is the ability of loan officers to game the technology by entering the appropriate data to gain a desired outcome.  However, I would argue this is not a technology issue; rather it’s a process issue. If underwriters must verify all sources of documentation and data used to render the lending decision this risk is eliminated or at least mitigated.  No system is immune from errant results if humans enter faulty data – whether accidentally or on purposefully.

Next, lenders depending solely on technology decisions soon find that they are at a loss to explain how a lending decision was made.  They suggest a complete lack of transparency exists in their transactions. This is certainly possible – if a lender abandoned sound credit policy, compliance, and other procedures in favor of accepting blind recommendations. It’s also likely they lacked the process, data and controls necessary to provide the level of transparency requested - or required - by investors, regulators, and securities holders.  Again, this is not a fault of the technology. In fact, an often ignored benefit of approaches based on business rules is the transparency it inherently provides compared to traditional implementation approaches.

Finally, some lenders complain about the inability of the technology to meet current investor requirements or that the technology is out of step with new regulatory requirements.  While non-automated solutions may ease these problems, it is more likely a case of outdated technology being relied upon.  It may also be the case of not making full use of the capabilities that exist in decision management platforms. It has been estimated that less than 60% of business rule management implementations (across all domains) truly leverage the full potential of these platforms and end up managing business rule projects like any other conventional software project (this will be the subject of a future essay). Clearly newer, more agile solutions facilitate keeping the systems constantly in sync with the market, competitors, customers, and regulators.

I believe the evidence is quite strong for the defendant in the case of The People vs. Business Decision Management. More importantly, what happens if this trend away from automated underwriting and decision making continues? Without correcting the fundamental flaws in process and control, we’ll see slower, inefficient processing, increased costs (more human-centric processing means more people and that means higher costs), and inconsistent, often poor decision-making. This is not the way to rescue an industry in turmoil. A 2011 Gartner BPM Survey indicated the top 5 business results achieved by organizations implementing BPM were:

• Reduced costs
• Improved quality of business processes
• Continuous process improvement
• Improved customer satisfaction
• Improved business process agility

Not coincidentally, these are exactly the features supposedly lacking in the systems accused of bringing down the mortgage industry. And they are exactly why business decision management needs to be the focus moving forward!

The sentence is clear: fix the core processes and make use of the full capabilities of these technologies. Technology is not the problem – it remains very much the solution!