Returns surge planning for January

"Last January plus a bit" is the usual method and it is wrong whenever the trading year was not identical. The inputs you need are ones you already hold: units shipped by category through November and December, your own historical return rate by category, and the lag between delivery and a return request.

Multiply through, then place the result on a calendar. The lag matters more than the rate, because it decides whether the work arrives as a manageable slope or as one wall in the second week. Orders delivered in late November come back first; gift orders come back later, sometimes considerably later if you extended the window, and they come back from people who were not the buyer.

Do it by category rather than in aggregate. Apparel and footwear behave nothing like homeware, and a peak that was unusually heavy on the high-return categories produces a January that looks nothing like the previous one even at an identical overall rate. The output you want is a weekly units-inbound profile for the first six weeks of the year, with a high case. That single chart is what the warehouse conversation is about.

Shopify's return rules are set on the store and, as the documentation states plainly, changes apply only to future orders so that customers who have already ordered are not affected. There is no retrospective switch. Whatever window was in force when the gift orders were placed is the window you are operating in January.

The rules cover more than the length. Shopify documents a return window — 14, 30 or 90 days, unlimited, or a custom period — and lets you choose whether it starts when an item is delivered or when the last item in the order is delivered, which is a meaningful difference on a multi-item gift order shipped in parts. You can set who pays for return shipping (free, a flat-rate fee, or the customer buying their own label), apply an optional restocking fee, and mark specific products or collections as final sale. Self-serve returns and cancellations have to be turned on for customers to raise them from their accounts.

So the October decision list is short and consequential: how long, measured from when, paid for by whom, with what excluded, and self-serve on or off. Get those five written down and most of January is administration rather than negotiation.

Customers are not the constraint. They will post the parcels back at a rate you cannot influence much. The constraint is what happens on your side of the door: booking the parcel in, inspecting and grading it, deciding its disposition, repacking or sending to disposal, and putting the sellable unit back into stock.

In any warehouse — yours or a 3PL's — that work sits behind outbound picking in the priority order, because shipping new orders is what generates revenue. Through a January with a residual sale still running, returns receiving is the queue that absorbs every delay elsewhere. The symptom is a backlog measured in totes and a refund SLA quietly slipping from three days to twelve.

This is where the money is. A unit received and re-listed in week one can sell at the price the original buyer paid. The same unit re-listed in week five arrives on the shop floor after the markdown, and the difference between those two outcomes is entirely within your control. Track days-to-restock as an operational metric with a target, because return rate is a merchandising number while days-to-restock is the one your January margin actually depends on.

January pairs the year's lowest revenue with the year's largest refund outflow, and the payment fees on the original transaction are generally not recovered. Finance teams that have never modelled this discover it as an unpleasant surprise in the second week.

Put the returns forecast into the cash forecast, using the high case rather than the expected one. The question to answer is not "what will returns cost" but "on which day is the gap between refunds out and revenue in at its widest", because that is the date that determines whether anything needs arranging in advance.

Two policy levers reduce the outflow without reducing the customer experience, and both are free. Offer an exchange before a refund, since a same-value exchange is an inventory movement rather than a payment. And offer store credit as a visible option, ideally with the arithmetic in the customer's favour, so that choosing it feels like a benefit rather than a consolation. Neither requires software; both require the returns flow to present them first, and the policy to say so clearly.

"Where is my refund" is the January equivalent of "where is my order" and it is generated by the same thing: silence during a period the customer cannot see into. It is entirely preventable with three messages you write once.

Tell them what happens and when, at the point they request the return: how long the post takes, how long processing takes at the warehouse, and how long their bank takes after you issue the refund — that last delay is not yours, and customers routinely attribute it to you because nobody told them. Then confirm when the parcel is received, and again when the refund is issued. Three automated messages remove most of the queue.

Be honest in them about January specifically. A store that says "returns are taking up to ten working days to process this month" gets fewer tickets than one that publishes a three-day promise it is not meeting, and it keeps the customers it would otherwise lose to the gap between the promise and the reality.

Buying a returns platform in December. It is an integration, a policy migration and a change to the customer-facing flow, landing in the month you least want a new failure mode. If the arithmetic supports buying one — and for a fair number of stores it does not — implement it in the spring and let it run for three quiet quarters before it meets a January.

Free returns adopted as a permanent policy without the arithmetic. It does lift conversion and it does increase return rates, and whether the trade is positive depends on your margin, your average basket and your category. Shopify supports charging a flat-rate return fee or having the customer buy their own label; those exist because free is not always the right answer. Work it out per category rather than site-wide, because the answer genuinely differs between a £30 accessory and a £200 coat.

And an unlimited returns window adopted as a trust signal. It moves your volume to a month you cannot predict, it makes forecasting impossible, and it guarantees that a share of the units arrive back after they can be resold at anything like full price. A generous, dated window achieves nearly all of the same reassurance and remains a thing you can plan around.

• Shopify support and store management
• Choosing a returns platform, and when Shopify is enough
• Returns data is product data