The Customer Wants a Refund. Your Staff Don't Know the Law.

The Real Problem

A customer walks into your clothing store in Newmarket with a dress she bought two weeks ago. She says the stitching on the hem came undone after one wash. She wants a full refund. Your staff member behind the counter isn't sure what to do.

Is this a manufacturing fault? Does the customer have to accept a repair first, or can she demand a refund? What if she washed it on the wrong cycle? Does it matter that she doesn't have the receipt?

Under the NZ Consumer Guarantees Act (CGA), goods must be of acceptable quality — fit for purpose, free from defects, durable, and safe. If a product fails to meet these guarantees, the customer is entitled to a remedy: repair, replacement, or refund, depending on whether the failure is minor or major. A hem coming undone after one wash on a $120 dress is likely a fault — and the customer probably has the right to choose between replacement and refund.

But your staff member doesn't know that. So one of three things happens:

1. They refuse the refund — "Sorry, we don't do refunds, only exchanges." The customer gets angry, threatens to complain to the Commerce Commission, and posts a bad review.
2. They give an immediate full refund — even though a repair might have been appropriate, costing you unnecessary money.
3. They call you — the owner — interrupting whatever you're doing to make a decision over the phone with half the context.

This happens every week. NZ retailers handle returns inconsistently because there's no standard process, no staff training on CGA obligations, and no documentation trail. When a complaint escalates, you have no record of what was said, offered, or agreed.

Why Existing Tools Don't Solve This

Shopify handles online returns as a transactional workflow — issue a return label, process the refund. But it doesn't guide staff on whether a refund is legally required, or whether a repair is the appropriate first step. It's a processing tool, not a compliance tool.

Lightspeed (Vend), the NZ-founded POS system, tracks returns as inventory transactions but offers no decision-support for CGA compliance. The return is recorded, but the reasoning isn't.

No NZ-specific returns automation tool exists for small retail. Large chains have legal teams and training programmes. Independent retailers have a printed returns policy taped to the counter — if they have one at all.

How AI Solves This

Your staff member photographs the damaged dress and sends the details to the store's AI assistant on WhatsApp:

Staff: "Customer bought this dress 2 weeks ago, $120. Hem stitching came undone after one normal wash. She wants a refund. Has the bank statement as proof of purchase but no receipt."

AI: "This looks like a product failing the Consumer Guarantees Act's acceptable quality standard — stitching shouldn't fail after one normal wash on a $120 garment. Here's the recommended approach:

CGA assessment: Likely a manufacturing defect (minor failure that can be remedied). The customer doesn't need the original receipt — a bank statement is acceptable proof of purchase under the CGA.

Recommended options to offer the customer:

1. Replacement — same dress, same size, if in stock
2. Refund — full purchase price to original payment method
3. Repair — you could offer to have the hem professionally repaired, but the customer is not obligated to accept a repair if they prefer replacement or refund for this type of fault

Suggested response: 'I'm sorry about the stitching — that definitely shouldn't happen. We can offer you a replacement or a full refund, whichever you'd prefer. Your bank statement is fine as proof of purchase.'

Document this: Record the customer's name, date, product, fault description, and outcome chosen."

The staff member handles the return confidently, the customer leaves satisfied, and the interaction is documented.

Building a Returns Knowledge Base

Over time, the AI helps you build a consistent returns process:

• RETURNS-POLICY.md — your store's returns policy, aligned with CGA requirements
• RETURNS-LOG.md — documented returns with fault type, assessment, and outcome
• CGA-GUIDE.md — plain-language summary of your obligations for common scenarios (faulty goods, change of mind, sale items, perishables)

When a new situation arises, the AI references these files for consistency. If a similar dress had a stitching issue last month and you offered a replacement, the AI ensures the same standard applies this time.

The Result

• Consistent returns handling — every customer gets the same fair treatment regardless of which staff member is working
• CGA compliance — staff know the legal obligations without needing a law degree
• Fewer escalations — customers handled well at the counter rarely escalate to formal complaints
• Documentation trail — every return is recorded with reasoning, protecting you if a dispute arises
• Staff confidence — junior staff handle returns without calling you for every decision

What AI Can't Do Here

• AI won't provide legal advice — for complex disputes or Commerce Commission complaints, you need a lawyer
• AI won't override your judgement — it provides guidance, you or your staff make the final call
• AI can't assess physical product quality remotely — staff still need to inspect the item
• AI won't handle truly abusive customers — those situations need a human response
• AI guidance is based on general CGA principles — edge cases may need professional legal input

Who This Is For

• Retail store owners whose staff handle returns inconsistently or incorrectly
• Clothing, homeware, and specialty retailers dealing with product quality complaints
• Any NZ retailer who wants to be CGA-compliant without memorising consumer law
• Stores with junior or casual staff who need clear guidance on returns decisions