Automate returns and exchanges for UK Shopify stores with n8n

To automate returns for a UK Shopify store, let the customer start the request, check the order against your policy and the 14-day cooling-off rules, send a return label when it qualifies, and keep refund approval with a person.

Why UK returns eat so much time

Returns tickets stack up for a simple reason: every one needs a policy check, a label, a status update and often a refund decision. On a quiet Tuesday that's fine. After Black Friday and through January gift returns, the same steps feel endless.

Labels aren't a one-click Shopify feature

UK merchants often say they can't buy return labels inside Shopify the way some US apps assume. Shop owners on forums talk about stitching Royal Mail Click & Drop, Parcel2Go or a self-serve returns app onto the store instead. Whatever you use, the workflow still has to decide whether a label should go out before anyone creates one.

January and post-Black Friday surge

Gift returns and "wrong size" messages spike after peak season. I covered the support questions that pile up in my Black Friday customer service checklist. This guide is the deeper how-to for the returns half of that list.

Two kinds of return (don't mix them up)

If your automation treats every return the same, you'll either refuse something you must accept or refund something you shouldn't without a look.

Change of mind / cooling-off

Under the Consumer Contracts rules for distance selling, the normal cancellation period for goods ends 14 days after the customer (or someone they name) gets the goods. They don't need a reason. GOV.UK also says that if you never told them about the right to cancel, they can cancel any time in the next 12 months until you tell them, then they get 14 days from that notice. Citizens Advice puts the extended window as up to a year and 14 days when the required cancellation information wasn't given in writing.

Some items sit outside change-of-mind cooling-off: personalised or custom-made goods, perishables, unwrapped CDs/DVDs/software, and sealed hygiene items once opened. Those still need a path for faults.

Faulty, not as described, or doesn't work

That's a different branch. GOV.UK is clear you must offer a full refund when an item is faulty, not as described or doesn't do what it's supposed to. Photos, a human review and sometimes a repair or replacement step belong here. Don't let a cooling-off template auto-approve every "broken on arrival" claim without a look.

Store policy vs legal minimum

Your published policy can be more generous than the law. Many UK shops offer 30 days as goodwill. The returns request demo on the demos page uses a 30-day demo policy with sample data. Say in the bot and emails which window you're applying, and keep faulty cases on their own track.

What to automate vs what to keep human

That split matches how I describe returns and refunds automation: the customer starts it, the checks run, and you approve the refund.

The n8n returns workflow (step by step)

Here's a practical shape you can build in n8n. Adjust the nodes to the apps you already pay for.

1. Trigger. A WhatsApp or website chat message, a short return form, or a Shopify return request. A WhatsApp AI agent can collect the order number and reason before n8n takes over.
2. Look up the order in Shopify Admin (line items, delivery date, tags, refund history).
3. Policy checks. Delivery date vs your window; product tags (personalised? hygiene-sealed?); reason code (change of mind vs faulty). Fail closed: if a check is unclear, queue a person.
4. Create the return record and message the customer with a reference and what happens next.
5. Label. When the request qualifies, create a label in the tool your warehouse already uses, then send the link. Use your existing carrier process (for example Royal Mail Click & Drop if that's what you run). Don't invent a new carrier API in the workflow until that account is connected and tested.
6. Notify the team in Sheets, Slack or email when a refund or exchange decision is needed.
7. Log every outcome (approved path, refused with reason, waiting on photo, handed to human) so you can audit a complaint later.

Exchanges instead of refunds

Where your policy allows, offer a size or colour swap before a refund. The automation can show in-stock alternatives from the same product family and only open a refund task if the shopper declines. Keep the tone helpful, not pushy. I'm not going to invent revenue-lift numbers for this; the win is fewer refund tickets and a cleaner stock move.

App vs custom n8n build

Pick the path that matches how customers already contact you. If you want one backend for returns, order updates and cart replies, custom n8n under AI automation for e-commerce is usually cleaner.

See it working

Watch the returns request demo: a fictional store, sample data, simulated chat. The shopper reports a damaged candle, the agent asks for the order number and a photo, checks a 30-day demo policy, creates a return with a label link, and offers a person on the team. Refund approval stays with you.

For the tracking side of post-purchase messages (not returns), see where's my order automation. For how returns show up inside a broader WhatsApp agent, read WhatsApp AI agent for Shopify (UK).

Want this mapped to your policy and stack? Book a free 15-minute audit.

Frequently asked questions