Mortgage Rework and Redesign Options
A mortgage lender processes 240 applications per week with 5 full-time underwriters working 40 hours each. End-to-end lead time averages 14 working days; hands-on value-added work totals 3.5 hours per application. A diagnostic found 9 handoffs, 3 systems with no integration, and 27% rework caused mainly by missing documents (61% of rework) and income calculation errors (23%). Staff re-key 41% of applicant data from the CRM into the core system. Two options are proposed. Option A: integrate CRM and core system and standardize a document checklist; cost EUR 180,000; projected rework 9%, lead time 8 days. Option B: create a central intake team that checks documents and re-keys data; cost EUR 45,000; projected rework 15%, lead time 11 days. The CIO favors A, the operations director favors B because it is cheaper and faster to implement. The board wants a fact-based recommendation.
1. Using process analysis, identify the dominant waste and its root cause(s). How should the two options be evaluated beyond their headline cost and lead-time figures?
Read the worked answer
The dominant waste is waiting, not touch labor. Value-added work is 3.5 hours against a 14-day lead time; at 8 working hours per day, that is roughly 112 hours of elapsed time for 3.5 hours of work, so flow efficiency is about 3%. The largest controllable driver is rework: 27% of applications, with 61% of that caused by missing documents. Re-keying 41% of data is a second waste - defects and unnecessary motion between systems.
Option A attacks both root causes: integration removes re-keying, and a standard checklist addresses missing documents. Option B adds a check after the defect rather than eliminating its cause; it may reduce rework but adds a handoff and does not fix the CRM/core disconnect. Evaluation should compare total cost of ownership, not just implementation cost: rework labor, underwriter capacity, error risk, customer wait time, and benefits ramp-up. Compute cost per application and payback period. Also check whether Option A's projected 9% rework and 8-day lead time are supported by a pilot or vendor evidence; otherwise treat them as assumptions to validate.
2. Recommend an option and support it with the data, including a capacity or flow argument. What risks and measurement plan would you attach?
Read the worked answer
Recommend Option A, but stage it. Capacity: 240 applications per week with 5 underwriters at 40 hours = 200 available hours. At 27% rework, effective work is inflated; reducing rework to 9% releases roughly 18% of application volume from rework, about 43 applications per week, which can be absorbed as reduced backlog or faster turnaround. Lead time improvement from 14 to 8 days also reduces work-in-process under Little's Law: WIP is approximately throughput multiplied by lead time. At 240 per week and 14 days (2.8 weeks), WIP is about 672 applications; at 8 days (1.6 weeks), it is about 384 - a 288-application reduction in exposure.
Option B's lower cost is attractive, but it leaves the integration defect and adds a central intake queue, which may become a new bottleneck. Risks for A include integration scope creep, data mapping errors, and underwriter change resistance. Pilot on one product line, measure rework by cause, straight-through rate, lead time at the 50th and 90th percentiles, and cost per application before scaling.