Returns Disposition Recommendation

It produces a prompt that asks AI to recommend one returns disposition action: restock, refurbish, liquidate, or scrap. The prompt is built around the item condition grade, return reason code, and estimated recovery value, so the output is tied to operational inputs rather than vague judgment. It is meant to standardize a decision, not replace warehouse or quality review where human approval is required.

Use it when your team handles repeatable returns decisions and needs a consistent first-pass recommendation. It is especially useful when the same product types come back often and the decision depends on a small set of variables. Do not use it as the only decision-maker for safety-critical, regulated, or high-value items that require inspection sign-off.

It can be used for every return that enters a disposition queue, or only for items that fall below a clear auto-restock threshold. Many teams run it after inspection data is captured and before the item is routed to inventory, repair, resale, or disposal. The key is to use it at the same point in the workflow each time so the recommendation is comparable.

A returns specialist, warehouse lead, QA associate, or operations analyst can run it, depending on your process. The person using it should understand your condition grading rules and what each return reason code means in practice. If your team has approval gates, the prompt should support the reviewer rather than authorize the final action.

The prompt should include constraints that prevent AI from recommending restock when policy, hygiene, warranty, or regulatory rules prohibit it. For categories like electronics, cosmetics, medical supplies, or food-contact items, the recommendation must defer to your internal compliance policy. In those cases, the template should be customized to require a policy check before any disposition output.

The biggest mistake is giving the model too little context, such as a condition grade without the return reason or recovery value. Another common issue is failing to define what each disposition means in your operation, which leads to inconsistent recommendations. Teams also forget to specify whether the output should include a short rationale, confidence level, or escalation flag.

Yes, the prompt can be paired with structured inputs from inventory, ERP, WMS, or returns management systems. The most useful setup is to pass the item ID, condition grade, reason code, and recovery estimate into the prompt, then write the recommendation back to the case record. If you integrate it, keep the output format strict so downstream automation can parse it reliably.

Ad hoc decisions are faster in the moment, but they are harder to audit and often vary by person or shift. This template creates a repeatable decision frame, which makes it easier to compare outcomes, train new staff, and spot policy drift. It also helps AI act as an assistant by recommending a disposition, while humans keep control of the final action.