Returns and Exchanges That Do Not Destroy Your Margin

A return costs you far more than the refund. Forward shipping is already spent, reverse pickup is a second freight charge, the item comes back needing inspection, and a share of what comes back cannot be resold at full price. In India, where COD is still common and RTO is a permanent line item, the difference between a well-run returns process and a default one shows up directly in margin. Most of that difference is operational, not technical.

What Shopify gives you natively

More than many merchants realise, which matters because a lot of stores buy an app before checking. Shopify's returns documentation covers three distinct actions that people often conflate:

• Refund. Returning payment, in full or in part, with no requirement that anything physically comes back.
• Return. Receiving the item back, optionally with a shipping label and tracking.
• Exchange. Sending an alternative item, typically a different size or colour.

Customers can request returns themselves from the order status page for delivered items, and cancel unfulfilled orders. You can configure return and cancellation rules that set which items are eligible and within what timeframe.

Return label availability varies by region, so verify what is actually available to you in India before designing a flow that assumes it. That check takes ten minutes and prevents designing around a feature you cannot use.

The number that actually matters

Most brands track return rate. On its own it is close to useless, because it treats a size exchange and a refund as the same event when their economics are opposite.

Track three things instead. Your refund rate, where money leaves. Your exchange rate, where the sale is retained. And your resaleable rate, meaning the proportion of returned stock that goes back to full price. A brand with a 20 percent return rate that is mostly exchanges and mostly resaleable is healthier than one at 12 percent that is all refunds on damaged goods.

Split those by reason and by product, and the picture usually gets uncomfortable quickly. Returns concentrate. A small number of SKUs generate most of them, and the cause is almost always the same three things.

Why things actually come back

Size and fit. Dominant in apparel and footwear, and largely a product page problem rather than a returns problem. A real size chart with measurements, not just S, M and L, plus a note on whether the fit runs small, prevents more returns than any policy change.

Not as described. Colour that photographs differently, material that feels cheaper than it looked, dimensions the customer misjudged. Fixable with better product data, which conveniently is the same work that helps you in search and with AI assistants.

Damage in transit. A packaging problem masquerading as a returns problem. If one SKU has a high damage rate, the answer is different packing, not a stricter policy.

Notice that the first two are solved on the product page. The most effective returns work usually happens before the order, not after it, and it is the same work our CRO audits keep recommending for entirely different reasons.

Push exchanges, not refunds

An exchange keeps the revenue and costs you one reverse leg plus one forward leg. A refund costs you the same logistics and gives the money back. So the design goal for the whole flow is simple: make exchanging easier than refunding.

In practice that means offering the exchange first in the self-serve flow rather than presenting refund as the default, making size exchanges genuinely frictionless because that is the most recoverable category, and considering store credit as a middle option where a customer does not want the same product. Credit at a small premium over the refund value converts a meaningful share of would-be refunds and is cheap to offer.

What does not work is making refunds deliberately difficult. It produces chargebacks, support load and reviews that cost more than the refunds did.

The COD and RTO overlap

In India this is where the real money sits. An RTO is not a return, it is a delivery that never happened, and it costs both freight legs with no revenue at any point. Brands with a serious RTO problem should fix that before optimising returns, because the numbers are usually larger. We covered the levers in our piece on reducing COD RTO.

The connection worth drawing is that both problems share a root cause: order intent that was never firm. A confirmed COD order and a well-informed purchase both come back less. Order confirmation over WhatsApp, honest delivery timelines and accurate product data reduce RTO and returns at the same time.

Write a policy you can actually run

The most common failure is a policy written for marketing that operations cannot execute. If your site says 30 day easy returns and your team takes two weeks to process a refund, you have created a support queue and a set of angry reviews.

State the window, the condition requirement, who pays return shipping, and how long a refund takes, in plain terms. Then make sure that text is the same everywhere it appears, including your policy page, your order confirmation, and your agents.md file if you have written one. An assistant quoting a return window your team does not honour creates a dispute you will lose.

The reverse leg is where the cost hides

Forward logistics get attention because they affect the customer experience. Reverse logistics get almost none, and that is where returns quietly become expensive.

Three things are worth measuring. How long a pickup actually takes from request to collection, because a customer waiting eight days for a pickup is a customer writing a review. What proportion of pickups fail on the first attempt, which in India is higher than most brands assume and doubles the cost of that leg. And how long returned stock sits before it is inspected and put back on sale, because stock in a corner of the warehouse is capital doing nothing during the weeks it would have sold best.

That last one is the most commonly ignored and often the largest. A brand with a fifteen day inspection backlog during festive season is holding sellable inventory out of the market at exactly the wrong time. Fixing it usually requires no software, just a decision about who inspects returns and when.

How long should a returns window be in India?

Seven to fifteen days is normal for most Indian D2C categories, with longer windows in apparel where sizing drives exchanges. Longer windows generally increase both returns and conversion, so this is a trade rather than a best practice. Pick one, measure it for a quarter, and adjust based on your own numbers rather than what a competitor advertises.

Should I offer free returns?

Only where the maths works. Free returns lift conversion and lift return rate at the same time, and which effect is larger depends on your average order value and margin. A common middle path is free exchanges and paid refund shipping, which nudges customers toward the outcome that keeps the sale.

Can customers request returns themselves on Shopify?

Yes. Shopify supports self-serve return requests from the order status page for delivered items, along with cancellation of unfulfilled orders, and you can set rules governing eligibility and timeframes. Turning this on removes a large amount of support email for most stores.

Do I need a returns app?

Not at first. Configure the native flow, run it for a quarter, and see where it actually strains. Stores usually need an app when they want automated reverse pickup with a specific courier, more complex rules than the native settings allow, or exchange logic tied to stock availability. Buying one before you know which of those you need means paying for the wrong thing.

What is the difference between a return and an RTO?

A return is a delivered order the customer sends back. An RTO is an order that never reached the customer, usually a refused or undeliverable COD parcel. RTO is generally the more expensive problem for Indian brands, and the two get conflated in reporting far too often.

Get your returns economics reviewed

Free review. Email hello@exactwhy.com with subject "Returns" and your return rate, your refund to exchange split, and your top five returned SKUs if you have them. We respond within 4 hours with where the cost actually sits, which is usually two or three products rather than your policy.

Paid work, Rs 25,000 to Rs 1 lakh. We set up the native returns flow properly, build the exchange-first path, fix the product page issues driving your worst SKUs, and put reporting in place that separates refunds, exchanges and RTO.

Ongoing Shopify development, Rs 20,000 to Rs 50,000 a month. Including the operational reporting that makes this visible month to month.

The brands that get this right are not the ones with the strictest policy. They are the ones who found the four products causing most of the problem and fixed the product pages.