Handling Returns and Exchanges over WhatsApp

A customer asking for a refund reads more carefully and waits less patiently than the one who bought. How that conversation goes decides not only one order but whether they order again. Automation should make the process faster and more transparent — not bury the customer in a form.

Anatomy of a Return Conversation

Almost every return needs the same five facts: which order, which item, the reason, the condition of the goods, and what the customer wants — refund, exchange or repair. The job of automation is to collect these without feeling like an interrogation. Asking six sequential questions exhausts people; pulling the order and letting them pick the item is far smoother.

Flow Design

• 1. Detect intent from phrases like refund, exchange, wrong item or damaged.
• 2. Verify the order by number or phone and show the purchased items.
• 3. Check eligibility from the system — return window and category rules, never the assistant's interpretation.
• 4. Record the reason and the requested outcome, with a free-text option.
• 5. Request photo evidence for damaged or incorrect items, collected in chat.
• 6. Start the process: generate the shipping label, send the address, state the expected timeline.
• 7. Update the status when the item reaches the warehouse and when the refund is issued.

Step seven is the most skipped and the most complaint-reducing. Silence during a return irritates customers more than the wait itself.

What Not to Automate

Anger detection should be simple but present: shouted capitals, repeated exclamation marks or references to formal complaints should stop the flow and bring in a person.

Turning Returns into Exchanges

Some returns are about unmet expectations rather than faulty goods, and a well-timed alternative keeps the revenue. The line to respect is insistence: offer one alternative without pausing the return — a larger size is in stock, would you prefer an exchange, otherwise the refund continues. The choice stays with the customer.

Return Data Is Product Feedback

• Return rate per product and the three most common reasons.
• Time from request to shipping label.
• Share of returns converted into exchanges.
• Share of return conversations escalated, and why.
• Share of customers who order again after a return.

That last metric is the important one: a well-handled return can build loyalty rather than end it.

Build the Rules In

Return rights and exceptions are set by consumer legislation and by your own distance sales terms. The eligibility check in the flow should point at that written policy and be updated whenever the policy changes. The assistant conveys your terms; it does not interpret the law.

With BozLat AI

The return flow is bound to the order record: eligibility comes from the system, photo evidence is collected in chat, and the customer is notified automatically as shipping and refund states change. See pricing for scope.