Synovus Reduced Underwriting Resubmits By 20% With GLSS

Tyson started by mapping his process. He created both a SIPOC and “As-is” Process Map to provide him with both a high-level and detailed view of the current underwriting process. He discovered that resubmits only occurred at two decision points at the end of the process, so he asked himself, “how many [resubmits] are getting here and can we stop [them] earlier?”

Tyson came up with a Data Collection Plan, measured their current process, and established baseline data for how their process was performing – finding 55% of underwritten loans were resubmitted for approval! Yikes!

Tyson wanted to determine where else waste lived in this process, so he performed a Value-Add Analysis for both the original underwriting and resubmission process. He found that the resubmission process contained 24% less value-add work than the first pass. What does this mean? Reduce resubmits!

The next for Tyson and his team was to determine the root cause of resubmits. Using a Fishbone Diagram, the team all agreed that analysts and bankers were the problem. Their hypothesis was that loans were resubmitted due to unfavorable results. However, when put to the Hypothesis Test, their theory fell flat. Their best guess was not a statistically significant root cause.

What do you do when all of your root causes prove to be false? You go back and look for more which is what Tyson did. The team went back to the fishbone diagram suggesting the next possible root cause, which did prove out.

The team turned their attention to the next most likely root cause of their high resubmittance rate: communication.

After surveying their analysts on the frequency of their communication with bankers, Tyson discovered that the less an analyst communicated with a banker, the more resubmits they received. The team had finally found their root cause.