Setting Disposition Rules for Returned Stock

A returned item lands on the receiving bench and someone has to decide, in under thirty seconds, what happens to it next. Restock it as new? Route it to a refurb queue? List it on a resale channel at a discount? Scrap it and eat the loss? Most merchants answer that question with a shrug and whatever the warehouse associate feels like doing that morning. The result is predictable: identical products get three different dispositions depending on who opened the box, margin leaks out of decisions nobody tracked, and the finance team has no idea why recovery rates swing 20 points month to month. Disposition is not a judgment call that should live in someone's head — it is a rules engine problem, and treating it as one is what separates operations that recover 70%+ of returned value from those stuck closer to the industry's uncomfortable average.

Why ad hoc disposition quietly destroys margin

Recommerce market data puts a hard number on the problem: only 48% of returned items resell at full price once they re-enter inventory. The other 52% split across markdown resale, refurbishment, liquidation, and outright write-off. That split is not fixed by nature — it is fixed by how fast and how consistently a retailer grades and routes each unit. A pair of shoes returned within 48 hours in original packaging should almost never end up in the same liquidation pallet as a pair returned after 40 days with a scuffed sole, yet without written rules, both frequently do, because the same overworked associate is triaging both under the same time pressure.

The fix starts upstream of disposition, at grading returned inventory, because the disposition decision is only as good as the grade feeding it. Garbage grade in, garbage routing out. Once grading is consistent, disposition rules can be applied mechanically — which is exactly what makes them automatable.

The four disposition lanes, defined precisely

Every returned unit should fall into exactly one of four lanes. Ambiguity between lanes is where value leaks, so the rule set needs to draw sharp lines rather than leave room for interpretation.

• Restock (A-grade): unworn, tags attached, original packaging intact, received within the return window with no odor, stain, or damage flags. Goes straight back to sellable inventory at full price.
• Refurbish (B-grade): minor cosmetic issues that a defined process can fix — re-steaming, re-boxing, button reattachment, cleaning. Goes to a refurb queue with a cost ceiling before it's allowed to become a restock unit.
• Resell / recommerce (C-grade): functional and clean but no longer first-quality — missing tags, opened packaging, a full return-window cycle passed. Routed to an outlet channel, open-box program, or third-party resale marketplace at a discount, as covered in recommerce and resale for returns.
• Liquidate or scrap (D-grade): damaged beyond economical repair, safety-flagged, counterfeit-suspected, or below the cost floor to process further. Goes to bulk liquidation or destruction, whichever nets more after handling cost.

Turning grader judgment into a rules table

The step-by-step mechanics of scoring condition at receiving are laid out in return grading and disposition steps; what matters at the policy layer is that each grade maps to exactly one disposition by default, with narrow, documented exceptions. A rules engine needs deterministic inputs: condition grade, days since delivery, category, unit cost, and channel-specific resale demand. Feed those four or five variables into a decision table and the disposition becomes a lookup, not a debate.

The override column matters as much as the default column. Rules without exceptions get gamed or ignored the first time an edge case appears; rules with two or three tightly scoped overrides per lane hold up because they've already accounted for the cases a human would otherwise escalate.

Where cost thresholds come from

Refurbishment thresholds should be set per category, not as a single blanket percentage. A $200 jacket can absorb a $15 dry-cleaning cost and still clear a healthy margin on restock; a $25 t-shirt cannot. Pull twelve months of refurb labor and materials cost against recovered resale value per category, and set the ceiling at the point where refurb spend stops paying for itself — typically somewhere between 15% and 25% of unit cost, but it genuinely varies by category and should never be guessed.

The moment we wrote down thresholds instead of trusting the floor team's gut, our refurb queue backlog dropped by a third — because half of what was sitting in it should have gone straight to liquidation on day one.

Automating disposition at receiving

A recovery platform earns its keep here by collapsing grading and disposition into a single scan-and-route motion at the receiving desk. The associate scans the item, answers a short condition checklist (or a computer-vision assist flags obvious wear), and the system applies the rules table instantly — printing a bin label for restock, refurb, resale, or liquidation without anyone needing to remember the current thresholds. That consistency is what turns disposition from a source of variance into a predictable, auditable process, and it's a big part of why platforms built for reverse logistics outperform spreadsheet-run pilot programs, a pattern McKinsey has documented broadly across supply chain digitization initiatives.

What breaks when disposition is left to instinct

Three failure modes show up over and over in operations that skip a formal rules table. The first is inconsistency across shifts: the morning crew restocks anything without a visible stain, the evening crew is stricter, and the resulting sellable-inventory quality varies by time of day, which customers eventually notice through complaint rates. The second is refurb queue bloat, where items sit for weeks because nobody set a clear cost ceiling, tying up working capital in stock that should have been liquidated on day one. The third, and most expensive, is silent over-liquidation — perfectly resellable B-grade items get lumped in with D-grade damage because grading it correctly takes ninety extra seconds nobody budgeted for, and those units are sold off in bulk at a fraction of their recoverable value. According to NRF research on returns handling, the labor cost of processing a single return can rival the margin on the original sale, which is exactly why every extra second of triage time has to be spent on the decisions that actually move recovered value, not spread evenly across all of them regardless of stakes.

Auditing the rules quarterly

Disposition rules decay if left untouched. Resale channel demand shifts, refurb labor costs change, and category mix moves with the catalog. Build a quarterly review into the process rather than treating the rules table as a one-time setup.

1. 1Pull actual disposition outcomes against the rules table and flag any category where overrides were used more than 15% of the time — that's a sign the default threshold is wrong, not that the exceptions are.
2. 2Compare recovered value per disposition lane against the prior quarter to catch resale channels that have softened.
3. 3Re-price refurb cost ceilings against current labor rates; a threshold set two years ago is almost certainly stale.
4. 4Sample-audit a batch of D-grade liquidations to confirm nothing that should have been refurbished slipped through on a busy day.

The bottom line

Getting from a 48% full-price resale rate to something meaningfully better isn't about grading harder — it's about removing the moment-to-moment discretion that lets identical items get different outcomes. A documented, threshold-based disposition table, enforced automatically at the point of receiving, is what makes recovery rates predictable enough to forecast and improve on, quarter over quarter.