The situation
Three related constraints, all of them capacity problems wearing different clothing.
Operational reporting consumed more than twenty hours of staff time each week, which meant the numbers reviewed in the operations meeting were several days stale by the time anyone saw them. Underwriters were spending a substantial part of their week on application intake rather than on credit decisions. Call center volume was being handled by whoever was available rather than by whoever was best equipped to resolve the issue.
The approach
Operational reporting. The reports themselves were not complex; the work was in the manual assembly. Automating extraction, standardizing the definitions so that the same metric meant the same thing across departments, and scheduling the refresh reduced the cycle from over twenty hours to under ten minutes.
Underwriting intake. Segmenting the application population identified the share that could be processed automatically with acceptable risk, which turned out to be around 65%. The genuinely careful work was defining the remaining 35% and the exception paths, because automation that quietly mishandles an edge case in credit underwriting produces losses rather than savings.
Call routing. Directing calls to associates with the relevant experience, and redirecting the resolvable cases to the website, reduced handling time and improved first-contact resolution without adding headcount.
The outcome
The weekly reporting cycle fell from more than twenty hours to under ten minutes. 65% of credit card applications moved through automated intake, returning underwriter capacity to decisions. Call handling time fell measurably through routing and web redirection.
What made it work
Involving the underwriters in defining what must never be automated. People protect processes they had no part in designing, and they improve processes they helped shape. The exception criteria they set were also better than the ones the project would have produced without them.