Retail Returns Practices That Protect Service and Margin

I'm John Mac, founder of Exit Evidence and OPENBATT. I help Shopify brands improve profitability and prepare for exit. I've walked the walk, having built and sold an omnichannel outdoor retailer.

Here's what I learned: not all orders are created equal.

If a customer orders the same jacket in two sizes, your order value doubles, but one jacket is coming back. Add outbound shipping, return shipping, time to inspect and repack, and the delay before it's back on sale. Nearly half of returns end up discounted and about 15% are a total loss. Once you add it all up, shipping and handling on those two jackets can reach $20-45.

So I track a different KPI: Average Kept Order Value (AKOV).

AOV is what rings up at checkout. AKOV is what's left after returns. I track it by SKU, size, acquisition channel and cohort. It has exposed many campaigns that looked like winners on the ad platform but were actually dragging down contribution margin.

I don't believe in checkbox return reasons.

Give customers a checklist and most will pick the first option just to move on. What customers do tells you more than what they say, and their purchase history shows whether something runs big or small. If you must ask, ask for specifics like shoulder width, sleeve length or waist size, and thank them with a small gift card.

Make returns easy on the front end and smart on the back end.

Let customers drop off returns at thousands of locations with a simple QR code. Behind the scenes, your business should already know where each item is headed.

Returns started within 3 days of delivery are 5-6x more likely to be in sellable condition than later ones. So: - Within 3 days: back on sale immediately - Within 15 days: inspected, then back on sale - After that: sold through subscription or open markets

For returns made at brick-and-mortar stores, decide where the item goes before accepting it: reorder, transfer or sale. Otherwise your stores fill up with SKUs that only sell online.

Factor reverse logistics into forward logistics, and you can cut new orders by 10-15%.

When I ran my own online stores, I approved every return the same way. Customer asks, label goes out, box comes back, and someone on my team figures out later what to do with it. That later decision is where I was losing money, because a scuffed item would sit in a pile until it was worth almost nothing.

So I moved the decision to the front. Every return request got a reason code before the label was issued, and every item that came back got graded the day it arrived. Sealed goods went back into inventory.

Opened but clean went out as open-box at a discount. Damaged went to a liquidation pile that we cleared on a regular schedule so it never aged.

For cheap items where return shipping would eat more than I could ever recover, I refunded the customer and told them to keep it or donate it. Buyers loved how easy that was, and I quit paying freight on stuff that was headed for the trash anyway.

The reason codes ended up doing the most work. I read them every week, and a cluster of 'not what I expected' complaints usually traced back to one product page with a misleading photo or a vague size chart. Fixing that page brought returns down on that product before they ever reached my warehouse.

1. My cheapest return is one I never have to make. All my time is spent pre-selling the return of a server plan (and like an opened jar of skin care, a returned server plan can't be resold). In the few cases where returns do happen, I treat returns as conversations rather than forms. It turns out most people who ask for a return of their money actually want their problem(s) with using the product to go away. So I offer to move them to a higher spec'd server, lower spec'd server, or even just add money to their account. This keeps the money on the books and is better for everyone than a refund. I tag the reason for return on each return and go through this with the marketing team on a monthly basis to look for trends and where our promises to sell our product and the actual product we're delivering have diverged.

2. I recall a circumstance where a client was wanting to withdraw from their Server Rental Agreement due to some confusion on our part. I was adamant that we stick to the terms & conditions, listing out each point one by one. But my senior colleague at the time took a different approach. He apologized, processed the refund immediately, and even sent the client a personal note. This client returned months later introducing us to two other clients. It just goes to show that in most cases, returns are actually a test of your ability to service your clients before an audience. I've since started to include returns in my budgets, almost as I would with any other form of marketing spend - in many cases, you actually end up earning more back than you've lost.

Our perspective comes from engineering consultancy, so I would approach this as a process improvement question. Customers need clear instructions, a straightforward way to raise a return and timely updates. Behind that experience, the business needs consistent checks to understand the product's condition and determine what value can still be recovered.

I would separate the customer service process from the product assessment. A customer should not have to repeat their explanation across several teams, while returned stock should move promptly towards an appropriate outcome, such as resale with an accurate condition description, repair or supplier review. Recording the reason for each return also helps identify recurring problems with packaging, product information or fulfilment.

The lesson I would emphasise is that recovering value starts before a return occurs. Clear specifications, accurate descriptions and checks before dispatch can prevent avoidable disappointment and reduce the work required afterwards.

Finding a balance for the challenges of product returns means moving from the warehouse-centric approach to focusing on the point when a customer first asks for a return to be processed. That is why I have found that the best approach to limiting losses for items that cannot be resold at their retail value is employing triage based on insightful documentation during my experience managing large customer service operations of well-known retailers. For that reason, we do not just provide the customer with a return label; instead, we process the documents in such a way that prompts the customer to fill out a detailed questionnaire with information about the state of a product and photo evidence. This way, we are able to immediately assess the situation. In cases when the returned product is of low value (when the costs of shipping and processing outweigh the recovery), it may be more profitable for the customer to keep the product and receive a refund. On the other hand, if the item is of high value but has been damaged, the system can route the return straight to the repair shop instead of the common fulfillment center thus avoiding additional losses during the logistics process.

A critical lesson that influenced my understanding of returns happened to me when I was responsible for running large BPO operations for various telecom and retail companies working on the US market. What we understood was that the return everything approach practically paralyzed our agents and the warehouse teams with the volume of low-value returned products that cost more than processing. Instead of providing convenience to our customers, such policy was putting an enormous strain on the company's logistics processes and thus delaying the reimbursement of funds due to the delays in processing. I learnt from that experience that convenience here is not about making everything returnable, it is about making it possible to provide an instant solution based on the item condition and using AI triage methodology to classify returns into those requiring immediate reimbursement, recycling, or repair. Thus, we keep our customers satisfied without having to subject our returned goods to retrading.

Keep the customer-facing return process straightforward, while tracking the costs behind each return by product and sales channel. Those costs can include reverse shipping, inspection, repackaging, fees, and any reduction in resale value. A key lesson is that gross margin alone can hide how much returns affect a product's actual contribution. Bringing those costs into the product economics can help teams spot return patterns and address their causes without making the process harder for customers.

Keep returns simple and fast for customers, while triaging every item on receipt: quick refunds for full-value resellables, detailed inspection for partial refunds, and routing unsellable items to repair, parts or controlled disposal. We rely on clear online product pages, accurate sizing and high-quality photos to cut avoidable returns before they happen, and on tight coordination between customer service, order prep and shipping to speed processing.

An early lesson was treating returns as product intelligence rather than just cost. After reviewing returned items we changed descriptions and sizing guidance, which reduced repeat issues and made returns easier to manage across our international e-commerce operations.

The early lesson that changed how we handle returns was that speed without visibility creates hidden costs. A return may be easy for the customer but difficult for the operation when its condition, reason, and arrival date are unknown. We now collect these details at the first practical opportunity and use them to guide the next step. This reduces repeated handling and helps our teams understand what is coming back.

We also learned to treat return data as a way to prevent problems rather than simply record them. When the same reason appears repeatedly, we review the original promise, order process, and information customers received. This helps us find the cause and make improvements. We can then address recurring issues while keeping legitimate returns simple for customers.

The biggest loss in returns is usually not the markdown you take on the unit. It is the units that were refunded and never actually came back, and across the accounts we manage that gap goes unnoticed because the refund and the return sit in two different systems.

On Amazon you can see it if you look. The FBA Customer Returns report lists every returned unit with its disposition: sellable, defective, customer damaged, carrier damaged. Reconcile that against refunds. Where a customer was refunded and no unit ever arrived at the fulfillment center, Amazon's own policy is to reimburse the seller once the return window lapses, roughly 45 days after the refund. That does not always fire on its own.

The second gap is disposition. A unit comes back in perfectly resellable shape, gets graded customer damaged at the receiving dock, and lands in unsellable inventory. Now the brand has written down an item that was fine. The inventory ledger and the removal order detail report are where that shows up.

The early lesson that changed how we handle returns for clients: reconcile weekly, not quarterly. The filing windows on these discrepancies are counted in days, so a quarterly review usually means the money is already out of reach.

On convenience, keep the front end generous and fix the back end instead. An easy return policy is rarely what costs you. Unreconciled returns, wrong dispositions, and missed claim deadlines are.

Make the return easy to start but have it start with a conversation, because that is where you can still limit the loss. A short "Before you buy" section we place above the product description gives a 45-day window from the day the order is delivered and tells the shopper to contact us first, and we set the return up from there. The window is generous and shown before checkout, yet the first step is a message to us rather than a package arriving with no context, so the return is arranged before anything is sent back. The same section says the oil comes in amber glass and should be stored with the cap sealed, out of direct light and away from heat, so care instructions travel with the promise. If I had to boil it down: publish the policy where people decide to buy, and make the customer's first move a conversation rather than a shipment.

I'd keep returns convenient by asking customers to check an order on the day it arrives, because a problem raised before anything has been worn is far easier to settle fairly than one that surfaces weeks later. Our refund policy builds that in: it asks customers to inspect their order when it reaches them and to contact us straight away if an item is defective, damaged or not what they ordered, so we can look at the issue and put it right. With period pants (period underwear), that first look does two jobs: it gives the customer a clear moment to speak up, and it means the problem is raised while the pair is still exactly as we sent it. If I were advising someone opening a shop today, I'd tell them to decide early what counts as a problem on arrival and how quickly they'll reply to it, because that first reply tends to shape how the rest of the return goes.

In our experience, the costly returns aren't abuse. They're big items that were the wrong choice from the start.

When a 269 lb steel cabinet set or a full-size griddle comes back, it can rarely be resold at full price. The box has been opened, moving it leaves scuffs, and shipping it twice can eat the whole margin. Like most retailers that sell bulky goods, we have a restocking fee to cover part of that.

The early lesson was that a fee only feels fair if people know about it before they pay. Found out after the fact, a restocking fee feels like a penalty. Stated right next to the price, the same fee is just part of the deal, and people choose more carefully because of it.

So the real work is preventing returns before the order, not managing them after:

- Put weight, dimensions and power or fuel requirements near the top of the page, so nobody orders a 36-inch griddle for a balcony that fits a 22-inch. - Answer fit and setup questions by phone before the sale. - Ask buyers to check big items at delivery and note any damage with the driver. A freight issue is far easier to sort out on day one than as a return three weeks later.

Convenient returns aren't only about free returns. They're about customers not needing to return anything in the first place.

I'd keep the return itself simple for the shopper and do most of the loss control before the order is placed, because once gloves or a jacket have been worn, they can't honestly be sold again as new. With riding gear, the return I'd work hardest to prevent is the wrong size, so every product page should have a measuring guide right next to the size options. When the size is wrong anyway, tell the customer the quickest fix up front, which is often to send the original back and place a new order in the right size rather than wait on a swap. For stock that's already marked down, I'd make it final sale and say so plainly before checkout, so nobody finds out after the fact. If I had to pass one habit on to a new store, it's reading the reason behind every return and fixing the page it points to, whether that's a size chart that doesn't match how the item fits or a photo that makes the color look different from the real thing.

Avoidable losses are best avoided before the loss happens. Our policy of double-checking sizes, colors and artwork with our customers before we start production has saved us more than any other loss avoidance measure we've taken.

When a mistake is made though, it's important that we make it right. Since all of our items are custom, all sales are final. Free and easy returns are important to online shoppers, as 82% of online shoppers surveyed by NRF and Happy Returns called free returns a major consideration when making a purchase, and an estimated 19.3% of online sales were expected to be returned in 2025. Since we can't always fall back on easy returns to make things right, we have to step up when we make a mistake.

One customer, a new small business owner, recently left us a review about her order with us at Arklavo. While she was happy with her products overall and said she plans to order from us again as her business grows, the order did take longer than expected and there was a mix-up with some of the colors. Luckily we were able to send out the correct color for free, and she still left us a four-star review!