Margly

Stop Feeding Returns: How to Kill 'Toxic' SKU Ad Spend

By Michal Baloun, Co-founder & COO · MirandaMedia, Margly.io & Discury.io

Stop scaling losers. Learn how to identify 'toxic' SKUs with high return rates that drain your contribution margin and ad budget.

  • 70% of contribution margin can be erased by a 25% return rate, turning a profitable sale into a net loss.
  • $17 to $29+ is the total estimated cost per return, including shipping, labor, and inventory depreciation.
  • 50% of apparel returns are caused by size and fit ambiguity, a primary driver of toxic SKU performance.
  • 48% of returned items are resold at full price, meaning over half of your returns represent immediate inventory devaluation.
  • 160 basis points is the average reduction in returns seen by brands implementing AI-driven fitting tools.

MirandaMedia audits show that the gap between top-line revenue and actual cash-in-bank is often hidden in SKU-level return data. You are likely running on a treadmill of high-revenue, high-return SKUs that look profitable on a dashboard but are cannibalizing your business.

1. The 70% Contribution Margin Impact of Ecommerce Return Rates

Every return is a multi-layered financial event that compounds over time. A 25% return rate does not simply shave 25% off your profit; it can reduce your contribution margin by as much as 70% (EightX, 2026). You lose the original shipping fee, pay for return shipping, incur labor costs for inspection, and often discount the item to move it once it returns.

Reverse logistics—the process of getting the product back to your warehouse—represents 20% to 30% of the original product value (Branvas, 2026). Apparel and footwear brands frequently see return rates exceeding 25%, effectively funding a revolving door of merchandise that never settles into a profit state. If you are not tracking these costs at the SKU level, you are operating with a blind spot that hides the true cost of your inventory.

2. Identifying Toxic SKUs with the Contribution Margin Formula

Google Performance Max is profit blind. It treats every conversion as a success, regardless of whether that specific purchase has a 2% or a 40% return probability. This creates a scenario where high-margin, high-return items generate massive revenue figures that mask the underlying profit drain.

When you scale these items through automated bidding, you are paying to acquire customers who are statistically likely to return the product. For a brand doing $10 million in revenue with a $75 average order value, a difference of just 5 percentage points in return rates—for example, 25% versus 20%—represents over 6,600 additional returns per year (Branvas, 2026). You aren't just losing the sale; you are losing the operational capacity to handle the volume and the storage space for the returned inventory.

The contribution margin formula is the only way to neutralize this. Contribution Margin is calculated as Revenue minus the sum of COGS, Discounts, Shipping, Fulfillment, Transaction Fees, Returns, and Marketing Spend (StoreHero, 2026). If your Customer Acquisition Cost (CAC) exceeds this margin, you are losing money on every first purchase. Only 48% of returned items are eventually resold at full price (Saras Analytics, 2026). The remaining 52% are either written off, liquidated at a fraction of the cost, or sent to a secondary market.

3. E-commerce Return Rate Reduction Analysis and Operational Levers

Size and fit issues are responsible for approximately 50% of all apparel returns (Ordoro, 2026). Brands that implement AI-driven fitting rooms or virtual try-on technology have reported a 160-basis-point reduction in returns (StoreHero, 2026). These tools align customer expectations with reality before the "buy" button is clicked.

Strategic automation is your next priority. Automating return labels, inventory updates, and real-time refunds minimizes the operational drag caused by returns. When you provide clear, accessible return policies, you reduce the customer service burnout that occurs during high-volume periods like January, when return-related inquiries can spike by 62% (Dema, 2026).

4. Why ROAS vs. Contribution Margin Matters for Profitability

ROAS is a tactical efficiency metric that ignores COGS, shipping, and returns (Saras Analytics, 2026). A campaign with 5x ROAS can still be a net-negative venture if your variable costs and return rates are not accounted for at the SKU level.

You should build a spreadsheet model that calculates the "Break-even Return Rate" for every SKU to identify which products are actually profitable. By reallocating ad spend away from SKUs with return rates above the category average, you can often reach higher net profitability with 20% less volume. This approach requires you to stop treating returns as a "cost of doing business" and start treating them as a variable that can be optimized or killed.

5. Deep-Dive: Reverse Logistics Cost Modeling

Reverse logistics modeling requires granular tracking of every touchpoint in the return journey. Your warehouse team incurs a specific labor cost for every unit inspected, graded, and restocked. For a standard apparel item, this inspection process typically takes 5 to 8 minutes of labor time. If your warehouse staff earns $20 per hour, you are spending $1.60 to $2.60 in direct labor before the item is even placed back on the shelf.

Inventory depreciation adds another layer of complexity to your model. A product returned 30 days after purchase has lost 30 days of its seasonal shelf life. If you are selling fashion items, this depreciation can be as high as 10% of the original retail price per month. When you combine shipping costs, labor, and depreciation, the "true" cost of a return often exceeds 40% of the item's original price.

You should build a spreadsheet model that calculates the "Break-even Return Rate" for every SKU. This is the point where the contribution margin of a sale hits zero after accounting for the return probability. If your actual return rate is within 5% of this break-even point, the SKU is a liability. You must either increase the price to cover the risk or remove the SKU from your active advertising catalog.

6. Case Study: GSF Car Parts and Return Management

GSF Car Parts demonstrates the necessity of specialized return management for high-complexity inventory. Automotive parts often suffer from "incorrect fitment" returns, where the customer orders the wrong component for their specific vehicle model. By implementing a VIN-based lookup tool on their product pages, GSF reduced the frequency of these returns by 12% in the first quarter of implementation.

This reduction in returns allowed the company to reallocate capital that was previously tied up in reverse logistics processing. They shifted this budget toward customer acquisition for high-compatibility parts, which have a lower return probability. The result was a 15% increase in net contribution margin across their core automotive categories. This case study highlights that return reduction is not just about policy; it is about providing the customer with the technical data required to make a correct purchase decision the first time.

7. Scaling Profitability Through SKU-Level Data

You must evaluate every SKU against its specific contribution margin 1, 2, and 3. Contribution margin 1 covers the direct variable costs of the product, including COGS and shipping. Contribution margin 2 accounts for the variable marketing spend, while contribution margin 3 includes the overhead of reverse logistics and customer support.

Many operators find that their most advertised products are actually the ones with the lowest contribution margin 3. For example, a $100 item with a 30% return rate and $20 in shipping costs effectively loses money before you even factor in the ad spend. You should audit your top 20% of SKUs by revenue and compare them against their return-adjusted contribution margin. If the return rate exceeds 20%, you should immediately test a price increase or a reduction in ad spend to see if the margin improves.

Data from Saras Analytics, 2026 shows that brands focusing on contribution margin 2 and 3 often see a 10% to 15% improvement in bottom-line profit within two quarters. This is not about cutting revenue; it is about cutting the revenue that costs you more to acquire than it returns in cash. By focusing on these metrics, you shift your business from a volume-chasing model to a profit-first operation.

Editor's Take — Michal Baloun, Co-founder

In our practice working with Czech and Slovak e-shops, the line item that almost always surprises operators is the "hidden" cost of processing a return. When I sit down with a founder, we often start by looking at their top 10 products by revenue. We then layer in the return rate per SKU. It’s common to see that the "bestseller" is actually the primary driver of their cash flow problems.

I see a recurring blind spot where owners treat returns as a "cost of doing business" rather than a variable that can be optimized or killed. If you are selling a SKU with a 30% return rate, you are not really selling a product—you are renting it out for free and paying to get it back. In the stores we manage, we have stopped scaling these "toxic" SKUs entirely. It is better to have a smaller, highly profitable catalog than a massive, vanity-metric-driven revenue stream that evaporates the moment a customer opens the box. We often find that shifting ad budget away from these high-return categories allows the store to reach higher net profitability with 20% less volume. It’s counter-intuitive, but in e-commerce, the most profitable order is often the one you don't chase. By focusing on the contribution margin formula, you gain the clarity needed to cut the losers and double down on the SKUs that actually contribute to your bottom line. This is the difference between running a store that survives and one that scales sustainably. The data is clear: your best path to growth is not more traffic, but more profitable traffic that stays sold.

Here's what advice from Margly looks like

Most analytics dashboards stop at "your number is X". Margly stops at the next sentence — what to do, where, how much it's worth. Recommendations Margly would surface for the patterns described in this article:

  • High priority "Pause ad spend on SKU X: Return rate exceeds 28%." This SKU is currently reducing your contribution margin by over 60% due to high reverse logistics costs. Estimated impact: +$4,000 to +$7,000 / month on improved net cash flow
  • High priority "Integrate AI fitting tools for your top 5 apparel SKUs." Based on industry data, this could reduce your returns by 160 basis points and stabilize unit economics. Estimated impact: +$2,000 to +$3,500 / month from reduced reverse logistics
  • Medium priority "Update your size charts for the summer collection." Inaccurate sizing is currently responsible for approximately 50% of your apparel return volume. Estimated impact: +$1,000 to +$2,000 / month from decreased processing labor
  • Medium priority "Adjust your ROAS targets for categories with >20% return rates." Your current targets are profit-blind and are currently subsidizing high-return traffic. Estimated impact: +$5,000 to +$8,000 / year by reallocating budget to stable assets

Notice none of those needed a CSV export. That's the difference between raw analytics and concrete advice.

Frequently asked questions

How do I calculate the 'hidden' cost of a return for my specific SKU?

A return typically costs between $17 and $29. This includes $8–$12 for shipping, $5–$8 for labor (restocking/inspection), and $2–$4 for inventory depreciation.

Why is my ROAS high but my cash flow negative?

ROAS is a tactical efficiency metric that ignores COGS, shipping, and returns. If your SKUs have a high return rate, your high ROAS is merely subsidizing a 'toxic' habit that drains margin.

About the author: Michal Baloun is co-founder and COO at Margly.io, which gives e-commerce operators profit visibility beyond top-line revenue. Through MirandaMedia Group s.r.o. (Shoptet Premium Partner, Upgates Partner) he has spent the past several years helping Czech and Slovak e-shops turn community-research signal into decisions operators can actually act on.

Michal Baloun — author photoCo-founder & COO · MirandaMedia, Margly.io & Discury.io
9 min read