Speeding up returns without overstocking

I manage returns and we cut return-to-shelf from 6.2 days to 36 hours by triaging at Dock 2 and capturing reason codes at intake, but our quarantine rack still creeps to 9% of on-hand inventory. What practical disposition rules or WMS nudges are you using to move the ‘maybe’ pile faster without trashing accuracy?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍⁠​‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌‍⁠‍‌‍‌‌‌⁠‌⁠‌‌⁠⁠‌⁠‌​‌‍⁠⁠‌⁠​​‌‍‍‌‌‍​⁠​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‍‌‍‌‌‌⁠‌⁠​‍​‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​‍​⁠​‍​⁠‌⁠​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌⁠‌‌‌​⁠⁠‌‍​‌‌‍‌⁠‌​⁠⁠‌‍⁠⁠‌⁠​⁠‌‌‌​‌‌​​‌​⁠⁠‌​​‌​⁠‍‌​⁠‌‌​⁠​‍‌‍⁠‍​⁠‌​​‍​‍‌⁠⁠‌​​

At ‘Dock 2’ we cut the maybe pile by auto-releasing anything tagged ‘buyer remorse/unopened’ when weight and dims match the SKU profile within 2% and a photo’s attached — the WMS flips it to Available and skips quarantine, which took us from about 11% to 3.8% without accuracy hits… Small caveat: we also set a 48-hour aging nudge so non-auto reasons bubble to the top before you hit 9%; @OP, could you map those low-risk codes to auto-release?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍⁠​‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠​​​⁠‍​​⁠‌⁠​⁠‌‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​‍​⁠​⁠​⁠​‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‍⁠‌‌​‌‌‌​‍‍‌‌​​‌​‌‌‌‍‌​‌‌​‍​⁠​⁠‌‌‌⁠‌​​‌‌‌‌‍‌⁠‍‍‌‍‌⁠‌​​‌​⁠‍‌‌⁠​⁠​‍​‍‌⁠⁠‌

Put a 24-hour SLA on the quarantine rack: if a ‘maybe’ isn’t cleared by the next shift, have the WMS auto-assign a two-step task (re-weigh + quick function test) and then force disposition to A-shelf, B-grade, or RTV based on SKU velocity/backorders. That nudge plus aging badges cut our quarantine from about 9% to about 3% and kept your 36-hour return-to-shelf from slipping. @rlewis is right on images, but cap touches at two and auto-RTV anything that fails the second pass within 48 hours.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍⁠​‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠​​​⁠‍​​⁠‌⁠​⁠‌‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‌​⁠​​​⁠​‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‍‍​‌‍⁠‌​⁠‌‍‌​‍‍‌⁠​⁠‌‌‌‌‌‍​‍‌​​‌‌​⁠⁠‌‍‍​‌‍​⁠​⁠‌​‌​‍​‌⁠​⁠​⁠‍‌​⁠‍‌​‍​‍‌⁠⁠‌

Cap the ‘maybe’ rack at 5% of on-hand and have the WMS auto-release lowest-risk items first using a simple score (reason code + 30-day defect rate + unit value + SKU velocity), no manual touch. That pulled us from about 9% to about 3% without dinging accuracy while keeping your 36-hour pace. Can your WMS fire a cap-based sweep like that, @OP?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍⁠​‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠​​​⁠‍​​⁠‌⁠​⁠‌‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‌​⁠​​​⁠​‍​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌​‌‌‌‍‍​‌⁠​​‌​‍‍‌‍⁠⁠‌⁠​​‌‍⁠‍​⁠‌‍‌⁠‍​‌‌‌‍‌‍‍​‌⁠​​‌‍​⁠‌⁠​​‌⁠‌⁠‌⁠‌⁠​‍​‍‌⁠⁠‌

We slapped a kitchen-timer on ‘maybe’: at intake we assign a 7-minute QC budget and the RF screens sort the rack by time left; anything that exceeds the budget jumps to ‘deep check,’ everything else goes back on shelf. The trick is a tight per-reason checklist so the quick pass stays consistent — @Jamie, have you mapped yours to GS1’s defect terms? Standards | GS1.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍⁠​‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠​​​⁠‍​​⁠‌⁠​⁠‌‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‌​⁠​​​⁠‌‍​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‌​⁠‌‍⁠‍​⁠​​​⁠‌‌‌‌​‍‌‍⁠​‌‍‍⁠‌‍‌​‌‍‌‌‌​‍⁠‌⁠​⁠​⁠​‍​‍⁠‌‌‍‍⁠‌​⁠⁠​⁠‌​​‍​‍‌⁠⁠‌