Returns on marketplaces: benchmarks across 6 product categories, who pays, how to cut the rate

On marketplaces a return is a customer right written into the law. The seller cannot refuse it, but can cut the rate by working on the description, photos, packaging and product quality. Sellers ask: "is it normal that 30% come back?" The answer: it depends on the category.

Return benchmarks by category:

• Clothing and footwear: 25–40% (try-on is a given, sizing is subjective).
• Cosmetics and perfume: 5–10% (sealed packaging, return restrictions).
• Electronics: 8–15% (defects or a mismatch with expectations).
• Home and garden: 5–12%.
• Kids' products: 10–20%.
• FMCG (household goods, cleaning chemicals): 3–8%.

If your SKU returns run 50% above the category benchmark, that is a signal to dig into the causes.

This Support material is a step-by-step methodology for analyzing causes and cutting the return rate.

Context is in the parent materials:

• Pillar "Marketplace promotion" — where returns sit in the overall MPO system.
• Pillar "SEO and card audit" — description and photos as causes of returns.
• Support "Hidden deductions on marketplaces" — the economics of returns and deductions.

On FBO:

• "Did not fit" (no complaint about the product): the customer pays the reverse logistics (50–150 rubles is withheld from the customer). The seller only pays to put the unit back into the warehouse (if it can be resold) or to dispose of it.
• "Defective product" (with photo evidence): the seller pays the reverse logistics (50–150 rubles) and loses the value of the item (it is written off as non-returnable).
• "Does not match the description": the seller pays the same as for a defect, plus the risk of a "faulty card" penalty of 1500–10,000 rubles.

On FBS:

• The return travels back to the seller (logistics of 100–200 rubles at the seller's expense in case of a defect, at the customer's expense in case of "did not fit").
• The seller decides what to do with the unit (resell, dispose of, repack).

On rFBS (Ozon):

• The seller arranges the return themselves. Full responsibility sits with the seller.

Cause 1. The size did not fit (clothing, footwear, furniture). The most common cause. Tactic: an accurate size chart, photos on a model with the model's measurements stated, a video review that focuses on the fit.

Cause 2. The color in the photo differs from reality. Tactic: shoot in daylight without filters, use several backgrounds, and state explicitly that "slight differences are possible due to monitor settings".

Cause 3. The product is smaller than expected. The customer expected a "100 ml cream" and received 50 ml. Tactic: state the volume/weight explicitly in the title and on the first photo.

Cause 4. Quality below expectations. Typical in a cheap niche. Tactic: realistic photos with no flattering, an honest description of the materials, and clear positioning as "budget" vs "premium".

Cause 5. Damage in transit (crushed packaging, broken item). Tactic: reinforced packaging (bubble wrap, cardboard inserts), a "Fragile" label on the box. See also Support "Shipping and labeling".

Cause 6. The wrong variant (different color, different set). Tactic: clear variant labeling and correct barcode mapping in the seller account.

Cause 7. The customer changed their mind. Impossible to eliminate entirely, but it can be reduced by describing the expected value better (video review, use cases).

Cause 8. Duplicate order. The customer ordered twice by accident. Tactic: an integration with the marketplace that flags duplicates (available through third-party services).

Step 1. Export the returns. On WB it is "Analytics → Returns", on Ozon "Analytics → Product returns". Export to Excel for a 30–60 day window.

Step 2. Categorize the causes. The marketplace often provides a text comment from the customer or a cause code. Group them into the 8 causes above. The output is a chart: "60% of returns are sizing", "25% quality", "15% other".

Step 3. Deep-dive the top cause. If 60% is sizing, open every return with a customer photo (where available) and read the comment. Patterns usually show up: "runs one size small" means the size chart actually needs revising.

Step 4. Actions. Every top cause gets its own plan. Sizing → a new size chart plus photos on a model. Quality → revise the photos or change the supplier. Color → a new photo shoot.

Step 5. Measure the effect after 30 days. Once the card has been changed, check the return rate over the next 30 days. If it stayed at the same level, the hypothesis was wrong — keep looking.

Specific thresholds:

• 15–25% above the category benchmark: improve photos and description, shoot a new video review.
• 25–50% above the benchmark: replace the photos, review the supplier, check the batch.
• > 50% above the benchmark: pull the SKU from sale, sort out the batch, relaunch.

High returns hit a card three times over:

• Economics. Every return costs 50–1500 rubles plus the potential loss of the unit.
• Rating. Cards with high returns get demoted in the algorithm's ranking. On WB especially — the buyout metric (share of customers who bought against the share who ordered) is calculated at SKU level.
• Reviews. A "defective" return often comes with a negative review. That is the second wave of losses — reviews affect the conversion of the next customers.

That is why cutting the return rate by 5 pp can deliver 15–25% growth in net revenue through all three effects.

A return on FBS comes back to the seller. What to do:

• If the packaging is unopened and the item is intact — repack it (if the packaging was damaged in transit) and list it for sale again.
• If there is minor packaging damage — sell it through "Discounted goods" (available on both WB and Ozon) at a 10–30% discount.
• If the item has an actual defect — write it off, send it back to the supplier if the contract allows it, or dispose of it.
• If the item came back intact but the seasonal peak has already passed — hold it until the next season.

• Pillar "SEO and card audit" — description and photos as causes of returns.
• Support "Hidden deductions on marketplaces" — the economics of deductions on returns.
• Support "Shipping and labeling" — packaging as a way to cut returns.
• Support "SKU unit economics" — calculating real margin with returns factored in.
• Returns on Ozon — who pays under FBO/FBS, how claims are handled and what can be disputed.

Export 30 days of returns. Calculate the rate for every SKU. For SKUs running 50% above the category benchmark, open every return, read the cause and group them. On WB, the free forseller AI audit automatically analyzes returns across the whole portfolio and flags SKU candidates for review.

In the standard Trusty MPO audit this layer is automated and cross-referenced with the others (SEO, content, rating).

Related breakdowns: Returns on Wildberries: who pays.