GB560 Unit 7 Process Design Proposal Example

Refund on Arrival, Inspect the Expensive Ones: A Unit 7 Process Design Proposal for an Appliance Returns Center

[Student Name]

Purdue University Global

GB560: Designing, Improving, and Implementing Processes

Unit 7 Process Design Proposal

[Instructor Name]

[Date]

The seller, its returns center and every figure and estimate are composite and illustrative; the methods and sources cited are real.

The Design in One Paragraph

Returns are sorted at the dock into three paths by value and history. Ordinary returns are refunded when they arrive and graded afterward. Expensive returns are graded and checked by loss prevention before any refund. Everything else keeps the current grading-first sequence. Exceptions leave the main path at the bench into their own lane, owned by one named lead. The daily export and the next-morning keying disappear because grading writes directly to the refund system.

Future-State Map

The map uses the same five lanes and symbols as the Unit 3 current state map, so the two can be compared step by step, and a legend marks each removed step.

Dock receiving. The dock scan now reads the carrier tracking number and matches it to the return authorization automatically, replacing the barcode that scanned poorly. The scan assigns each return to a path.

Fast lane. Returns under $150 from accounts with no dispute history, about 71 percent of volume. The refund is issued at the dock scan and posts to the shopper's card at the payment processor's next daily settlement, usually within a day. The unit is then unpacked and graded as before, but grading no longer holds up the refund.

Inspection lane. Returns priced above $400, about 4 percent of volume. The unit is graded and checked by loss prevention before the refund posts, which meets requirement four from Unit 2.

Standard lane. Everything else, about 25 percent of volume, keeps the grading-first path, but the grade now triggers the refund directly.

Exception lane. Returns that cannot follow any lane without a decision outside the procedure leave at the bench into a drawn lane with five branches, one for each kind of exception counted in Unit 6. One named exception lead owns the lane and clears it every day.

Removed steps. Four boxes from the old map are gone: the internal label, the afternoon export hold, the export file itself and next-morning refund keying.

The fastest refund the company can offer is one that does not wait for anything, so the design lets most returns stop waiting at the dock.

Why Refund Before Inspection

Many retailers refund on carrier scan for low-value items and trusted customers, accepting a small loss rate in exchange for speed and fewer service contacts. The proposal prices that exposure rather than assuming it away. Fast-lane returns number about 53,500 a year with an average refund of about $71. If 0.35 percent turn out to be empty boxes or the wrong item, a rate based on loss prevention's case history for low-value returns, the added loss is about $13,300 a year.

Against that, the fast lane removes the wait that causes most refund-status contacts and chargebacks. If it removes 64 percent of refund-status contacts, about $54,200 a year, and nearly all chargebacks filed before a refund posts, about $27,100, the avoided cost is roughly $81,000 a year. The exposure is about a sixth of the saving. The 0.35 percent loss rate is the assumption most in need of testing, and the Unit 8 pilot sets a limit on it.

Why Grading Stays Manual

Grading a returned appliance involves judgments a rule handles badly: whether a scratch on an air fryer matters, whether a smell means use, whether a missing manual changes the grade. A rule-based grade would produce confident errors that reach the next buyer faster, adding to failure mode four. The design removes only steps that were pure waiting or copying, and leaves judgment where it is. Fraud review on expensive units stays with loss prevention for the same reason. Davenport (1993) argues that process redesign should use technology to remove handoffs and delays, not to replace the judgment that gives a process its value, and the proposal follows that line.

The Exception Lane

The lane has five branches. Returns without a usable authorization go to an order search, now much rarer because the tracking number match replaces the barcode. Incomplete bundles go to a shopper contact with a two-day reply window. Gift returns go to a purchaser lookup. Recalls follow the recall procedure. Marketplace items are routed to their seller. The exception lead clears the lane daily and reports any item older than 48 hours each morning. Hammer (1990) warned that exceptions handled informally consume the time the main path saves, and drawing the lane is the response.

Change Table

Each change is linked to the Unit 5 failure it targets.

Tracking-number match at the dock. Removes most of mode 2, returns without usable authorization.

Fast-lane refund at the dock scan. Removes most of mode 1, refund-status contacts, and mode 5, chargebacks before refund.

Grade writes directly to the refund system. Reduces mode 1 for the standard lane and removes the keying step where some wrong amounts, mode 3, began.

Inspection lane for units above $400. Keeps the check that catches mode 6, swapped parts.

Manual grading kept, with a regrade sample. Keeps the check that catches mode 4, defective units graded sellable.

Exception lane with an owner. Prevents exceptions from waiting without a decision.

Expected Effects

The proposal expects median dock-to-refund time to fall from 3.34 days to about one day, driven by the fast lane and the removal of the export and keying. Touch time should fall slightly, since the internal label and keying disappear, though grading time does not change. The Unit 2 refund-timing requirement should be met for the fast and standard lanes; the inspection lane may need a day longer.

What the Design Does Not Change

The bulk carton wait before unpacking, the largest single delay in Unit 4, is reduced by the dock sort but not removed, because unpacking still happens at a separate bench. For fast-lane returns it no longer delays the refund, so it matters less to shoppers. It still delays restocking, so the inventory planner's 48-hour requirement may remain hard to meet. That is noted as a risk for the pilot to measure.

Assumptions to Test

The 71 percent fast-lane share, the 0.35 percent loss rate and the 64 percent reduction in contacts are estimates. Each will be measured in the Unit 8 pilot. Spear and Bowen (1999) describe well-designed processes as built to test their own assumptions, and the pilot is where this one is tested.

References

Davenport, T. H. (1993). Process innovation: Reengineering work through information technology. Harvard Business School Press.

Hammer, M. (1990). Reengineering work: Don't automate, obliterate. Harvard Business Review, 68(4), 104-112.

Spear, S., & Bowen, H. K. (1999). Decoding the DNA of the Toyota Production System. Harvard Business Review, 77(5), 96-106.