Ecommerce Return Rate (2026 Benchmarks by Category, the CFO Formula, and How Returns Hit Contribution Margin)

Returns are the most under-counted variable cost in DTC. Most brands track the return rate as an ops metric (how many came back) and stop there. The CFO version goes further: returns reduce net revenue (compressing gross margin), they add processing costs that sit in CM2 (compressing contribution margin), and they affect the customer cohort curve that feeds LTV (compressing customer-level economics). This page is the operator's playbook for the full picture: the formula, vertical benchmarks from 35 DTC brands, the all-in cost of a return (it's not just the shipping reimbursement), the seven levers that actually reduce returns without hurting conversion, and how to size a returns reserve correctly. If you've been reporting a 5% return rate to investors because that's what the Shopify dashboard shows, this page is going to be uncomfortable. The real number is almost always higher and almost always more expensive than it looks.

What is the ecommerce return rate?

The ecommerce return rate is the percentage of units (or dollars) shipped that are returned by customers within a defined window. There are several flavors and brands often conflate them, which leads to numbers that look reassuring but aren't comparable across periods or vendors.

• Unit return rate: units returned divided by units shipped. The most common version.
• Dollar return rate: dollars refunded divided by gross sales. The CFO version, because it sizes the returns reserve.
• Gross return rate: all returns, including exchanges and store credits.
• Net return rate: only returns that result in a refund (i.e., lost revenue).
• Cohort-matched return rate: returns from this month's shipments / shipments from this month. This is the one you want.

The return rate formula

The basic formula:

Plain English: what came back divided by what went out.

Unit form: Unit Return Rate = Units Returned / Units Shipped

Dollar form: Dollar Return Rate = $ Refunded / Gross Sales

Cohort form (do this one): Cohort Return Rate = Units Returned from Period X Cohort / Units Shipped in Period X

The cohort-matching matters. Return lag is typically 14-30 days outside Q4 and 30-60 days in Q4. If you match May returns to May orders received (not shipped), you'll overstate Q1 returns and understate Q4 returns by 3-6 percentage points. Most ecommerce platforms default to calendar-period matching, which is wrong for management reporting. Fix it in your data warehouse.

Average ecommerce return rate (2026)

The blended cross-category number sits around 16-18% in 2026, down slightly from the post-COVID peak of 20%+ as brands have tightened policies. Three caveats on the blended number: (1) it's heavily skewed by apparel, which is roughly 40% of the ecommerce dollar volume and runs 2-3x the cross-category rate; (2) it varies by 10-15 percentage points across verticals; (3) it doesn't reflect Q4 spikes. For planning, use the vertical-specific number from the next section.

The trend that matters: the National Retail Federation (NRF) tracks online returns annually. The line is consistently up over the last six years:

Sources: NRF + Happy Returns annual reports. The number is now structural, not a pandemic artifact, and Q4 still spikes 4-6 percentage points above the annual blended rate.

Return rate by industry / category (2026 benchmarks)

Benchmarks from our 35-brand DTC portfolio, cross-referenced with NRF and Shopify industry data:

Apparel is the outlier — it's 12x supplements and 2-3x most other categories. Within apparel, the variance is also wide: subcategories with fit-dependent silhouettes (dresses, denim, formalwear) run 30-40%, while standardized basics run 8-12%. The category aggregate hides the variance; build the table by subcategory if you sell across multiple silhouettes.

Return rate by channel

Amazon is the most expensive return channel because the platform's customer-friendly default policy drives the highest return rate AND Amazon charges a return processing fee on top. If you sell on Amazon FBA at a 35-45% return rate vs DTC at 25%, you need a 4-6 point higher gross margin on the Amazon SKU to maintain CM2 parity. Most brands don't price for that and silently lose money on Amazon.

Returns and contribution margin: the CM2 hit

Returns hit two places in the contribution margin ladder. First, net revenue (the top line) shrinks by the dollar amount of refunds. Second, the variable cost of processing the return sits in the CM2 layer.

The right way to size the returns reserve per order:

Returns Reserve per Order = (Return Rate × Cost-to-Process per Return) + (Non-Resellable Rate × COGS)

For an apparel brand at 25% returns, $9 cost-to-process per return, 15% non-resellable, $24 COGS, $89 AOV:

• Cost-to-process component: 25% × $9 = $2.25 per order
• Non-resellable component: 25% × 15% × $24 = $0.90 per order
• Total returns reserve: $3.15 per order

Most brands book only the return shipping reimbursement ($1-1.50 per order on average), which understates the reserve by 50-70%. The shortfall doesn't show up until quarter-end when the reserve is trued up and gross margin moves negatively for "unexplained" reasons. For the full CM2 build, see our contribution margin guide and the CM2 glossary post.

The all-in cost of a return

The all-in cost has five components most brands undercount:

For supplements, the non-resellable write-off is nearly 100% of COGS because consumables can't be put back into inventory once opened. The return shipping reimbursement is the smallest component, even though it's the line founders see on their 3PL invoice and assume is the whole cost.

Returnless refunds: when they make sense

A returnless refund (sometimes called "keep it") refunds the customer without requiring them to ship the product back. The economic rule: offer a returnless refund when the all-in cost of processing the return (return shipping + inspection labor + write-off probability) exceeds the value of the recovered product.

The rough threshold: when unit cost (COGS) is under $20-25 and the item has low resale value when returned (consumables, low-margin basics, opened beauty), offer a returnless refund. Amazon has used this approach for years on low-cost items. Loop, Returnly, and Narvar all support returnless refund rules in their workflows.

The trade-off is policy abuse: a small percentage of customers will identify the threshold and exploit it. Cap returnless refunds per customer per year (3 in 12 months is a typical limit) and track abuse-flag patterns to prevent gaming.

Bracketing and serial returners

"Bracketing" is when a customer orders multiple sizes or colors with the intent to return all but one. Apparel-specific; it accounts for 30-40% of apparel returns on average. The customer behavior is rational (uncertainty about fit is high; the cost of buying two is low when returns are free), but it's expensive for the brand.

Levers to reduce bracketing without killing conversion:

• Fit confidence tools (True Fit, Fit Analytics, Bold Metrics) reduce the perceived uncertainty that drives bracketing
• Detailed sizing data on PDP — model height/size, garment measurements, fit notes from real customers
• Free exchanges vs paid returns — exchanges are cheaper to process than refunds and don't lose net revenue
• Return fee for orders with 2+ of the same SKU in different sizes — explicit bracketing deterrent

"Serial returners" are customers whose lifetime return rate is 2x or more the brand baseline. They account for typically 5-10% of customers but 20-30% of returns. Loop and Narvar both surface serial returner reports. Options range from gentle (educational fit content) to firm (gate returns after 3-4 in a 12-month window, or charge a restocking fee for high-return customers).

Wardrobing is the third behavior worth naming. A customer buys an item, wears it once (often with the tags tucked in), then returns it within the window. It's most common in formal apparel, dresses for events, and special-occasion footwear. Detection: photo-condition checks at intake, tamper-evident tags on premium SKUs, and patterns in the returns data (same customer + same SKU type + return-within-7-days). It's the hardest of the three behaviors to design against because the customer has plausible deniability; the response is policy (final-sale on event/formal wear) more than tech.

How to reduce return rate (8 levers)

1. Fit content and sizing tools. True Fit, Fit Analytics, Bold Metrics, AR try-on. Apparel returns drop 3-6 percentage points. Highest ROI lever.
2. Detailed PDPs. Multi-angle photos, video, customer try-on photos, material composition, care instructions. Across categories, reduces wrong-product returns 2-4 points.
3. Exchange-first flows. Loop, Returnly, Narvar default the return flow to "exchange" instead of "refund". 30-50% of would-be refunds convert to exchanges. Preserves revenue without forcing repurchase.
4. Return-fee gating for repeat returners. After 3+ returns in 12 months, charge $5-7 per return shipment. Doesn't affect first-time customers; cuts serial-returner volume.
5. Smaller, more curated assortments. Broader catalogs increase choice paralysis and bracketing. Curated drops (Aimé Leon Dore, Aritzia) consistently show 20-40% lower return rates than open-catalog peers.
6. Post-purchase comms. Sizing confirmation emails, fit tips specific to the SKU, "how to wear" video. Catches return triggers before the item ships back.
7. Tighten return window. 30 days instead of 60-90 cuts returns 2-3 points. Doesn't affect conversion materially when communicated clearly at checkout.
8. Address-verification at checkout. Wrong-address ships drive 5-8% of "lost in transit" returns. Tools like SmartyStreets, Loqate cut this to under 2%.

Most brands try one lever at a time and see modest results. Run 3-4 in parallel over a quarter and apparel return rates can move 4-8 percentage points, which is the difference between a 28% return rate and a 22% one. The math on that is enormous: at $89 AOV and $34 all-in cost per return, every percentage point of returns saved on $10M revenue is $382K of recovered cost.

Return rate on the income statement

Returns hit two places on a GAAP income statement. (1) Net revenue is gross revenue minus refunds and allowances; refunds reduce net revenue at the top line. (2) The returns reserve (the estimated future cost of returns on already-shipped orders) is booked as a contra-revenue or as a COGS line, depending on the accounting policy. Public DTC companies typically disclose the returns reserve in their 10-K footnotes; it usually runs 2-6% of gross revenue.

Two things to get right: book the reserve at the time of sale, not when the return arrives (otherwise your P&L lags reality by 30-60 days); and update the reserve assumption every quarter against actual returns received. Brands that under-reserve consistently get a margin haircut at year-end when the auditor trues up the reserve to the actual rate.

Holiday and Q4 return rate

Q4 return rates run 1.5-2x the trailing twelve-month baseline. The driver is gift-giving: items received in December are returned in January (or February for the late laggards), so December cohort returns extend into Q1. The typical pattern for an apparel brand: 25% TTM baseline → 35-45% on the November + December cohort. Plan cash and reverse-logistics capacity for the 1.5x bulge or you'll be paying premium 3PL rates in January to process the backlog.

The cash impact is more important than the operational one. A brand that does 35% of annual revenue in November + December at a 40% Q4 return rate has 14% of annual revenue (35% × 40%) being refunded in January, which is the same month that working capital needs are highest for spring inventory. Most cash-flow blowups in February-March for ecom brands trace back to underestimating Q4 returns. Model the cash, not just the rate.

Common return-rate mistakes

1. Calendar-period matching instead of cohort-period matching. Overstates Q1 returns, understates Q4 returns by 3-6 points.
2. Only counting refunds, not gross returns. Net return rate is right for revenue accounting but wrong for ops capacity planning.
3. Booking returns reserve only against shipping cost. Misses inspection labor, write-offs, and the largest component (lost CM2 if no reorder).
4. Blending channels. Amazon and DTC return rates can differ by 10-20 points; blending hides the channel-level economics.
5. Reporting category averages instead of subcategory. Dresses are 35-40%; basics are 10%. The aggregate apparel number doesn't help you act.
6. Treating returns as an ops problem, not a CFO problem. Returns hit revenue, GM, CM2, and LTV simultaneously. The fractional CFO version is to model returns as a variable cost line, not as a quarterly surprise.

Free returns + contribution margin calculator

The Eightx contribution margin calculator lets you set a return rate and cost-to-process per return and see the impact on CM2 directly. Open the calculator. For a full diagnostic of your returns impact on gross margin, CM2, and customer LTV, book a 30-minute call.

How returns connect to the rest of your P&L

Return rate never moves in isolation. A high return rate inflates your shipping cost as a percent of revenue because you pay freight twice, it changes the math behind free shipping, and it loads your support team, which shows up in customer service cost per order and tickets per 1,000 orders. It also distorts the top line, which is the gap between gross sales and net revenue. If you sell into Australia, benchmark against the Australian return rate by vertical rather than the US figure.

Conclusion

Returns are the most under-counted variable cost in DTC. Get the formula right (cohort-matched, not calendar-period). Get the benchmark right (apparel 25%, beauty 8%, supplements 4%, food 2% — by subcategory if you sell across silhouettes). Get the cost right (the all-in cost is $34-50 per apparel return, not the $5 shipping reimbursement on the 3PL invoice). Get the reserve right (book the full reserve, not just shipping). Then run the eight reduction levers in priority order; most apparel brands can move return rates 4-8 percentage points in a single quarter, which is worth more than any single marketing optimization at scale. The brands that handle returns as a CFO problem rather than an ops problem are the ones whose P&L doesn't blow up in January.

Frequently Asked Questions

what is the average return rate for ecommerce?

Blended ecommerce return rate sits around 16-18% across categories in 2026, which hides huge variance. Apparel runs 20-30% (median 25%); beauty 6-10% (median 8%); supplements 3-6% (median 4%); food and beverage 1-3% (median 2%); home and lifestyle 8-15% (median 12%); electronics 8-12% (median 10%). The blended number is almost useless for planning; use the category number for your vertical instead. Returns also spike in Q4: holiday return rates run 1.5-2x the trailing twelve-month baseline as gift-given items are sent back in January.

what is a normal return rate for apparel?

Apparel return rates run 20-30% for established DTC brands, with median around 25%. Subcategories vary: footwear 18-22%, denim 25-35%, dresses 30-40%, basics (tees, undergarments) 8-12%. The driver is fit. Brands with broad size ranges and limited fit guidance see the highest rates; brands with strong size data, fit quizzes, or fit-prediction tools (True Fit, Fit Analytics, Bold Metrics) cut returns 3-6 percentage points. Try-before-you-buy programs (Aerie, Stitch Fix, Wantable) intentionally accept higher returns in exchange for higher cart size and conversion.

how do you calculate return rate?

Two formulas, depending on what you're measuring. Unit return rate = Units Returned / Units Sold for a given period. Dollar return rate = $ Returned / Gross Sales for the same period. Most brands report the unit version because it tracks customer behavior; the dollar version is what your CFO uses to size the returns reserve on the income statement. Use the same cohort for both numerator and denominator: returns received in May from orders shipped in May, not May returns against May orders. The shipping-period match matters in Q4 where return lag is 30-60 days and a January-on-December match makes the rate look 2x worse than reality.

do returns count against gross margin or contribution margin?

Both, but in different ways. Returns reduce net revenue (which compresses gross margin from the top because the denominator shrinks). They also add a returns-processing cost that sits in the CM2 layer of contribution margin (return shipping reimbursement, restocking labor, repackaging, write-offs for unsellable returns). The CFO version is to book a returns reserve at the time of sale equal to (historical return rate × cost-to-process per return + non-resellable rate × COGS). For apparel at 25% returns with $9 cost-to-process and 15% non-resellable, that's $2.25 + (3.75% × $24 COGS) = $3.15 per order in returns reserve. Most brands undercount this because they only see the shipping reimbursement, not the labor and write-off.

how much does a return cost a DTC brand?

All-in cost of a single return for an apparel brand in 2026 runs $7-12: return shipping reimbursement $4-6, 3PL restocking and inspection labor $2-3, repackaging $0.50-1, and the probability-weighted write-off for items that come back unsellable (typically 10-20% of returns). Add the lost margin contribution from the order itself (if the customer doesn't reorder), and the true unit cost of a return is closer to $25-35 once you include lost CM2 on the original sale. Lower-cost categories (supplements, food) have cheaper returns but proportionally higher write-off rates because consumables can't be resold, so the all-in cost can be 100% of COGS plus the shipping.

what is a returnless refund and when should I offer one?

A returnless refund (sometimes called 'keep it') is when you refund the customer without requiring them to ship the product back. It's worth offering when the all-in cost of processing the return exceeds the value of the recovered product. The rule of thumb: if the unit cost (COGS) is under $20-25 and the item has low resale value when returned (consumables, low-margin basics, opened beauty), offer a returnless refund instead. This is standard practice at Amazon for low-cost items and is increasingly built into Loop, Returnly, and Narvar workflows. The trade-off is policy abuse; cap returnless refunds per customer per year to prevent gaming.

why is the apparel return rate so much higher than other categories?

Three reasons. First, fit. Sizing is inconsistent across brands and inherently uncertain online; customers buy two sizes intending to return one (bracketing), which alone accounts for 30-40% of apparel returns. Second, fashion. Apparel is bought emotionally and returned rationally after the in-person try-on, which adds 10-15 percentage points vs categories where the purchase intent is more functional. Third, return policy permissiveness. Apparel categories have historically accepted free returns as the cost of conversion; the elevated rate is partly a brand choice. Brands that have tightened policies (charging restocking fees, requiring return windows under 30 days, banning serial returners) have cut returns 4-8 percentage points without proportional conversion loss.

how does Amazon's return policy affect FBA gross margin?

Amazon's customer-friendly return policy means FBA brands typically see return rates 1.5-2x higher than the same SKU on DTC. Amazon charges a return processing fee ($1-5 per return depending on category) plus the cost of repackaging unsellable inventory. Customer-damaged returns are reimbursable through Amazon's reimbursement claims process but require active management; brands that don't file claims monthly leave 1-2% of revenue on the table. The net effect is that Amazon gross margin (after factoring in higher returns and reimbursement claims) typically lands 3-6 percentage points below the same SKU on DTC, even before the referral fee and FBA fulfillment fees.

what's the difference between gross and net return rate?

Gross return rate counts every return: refunds, exchanges, and store credits. Net return rate counts only the returns that resulted in a refund (i.e., lost revenue). For brands with an exchange-first flow or active store-credit incentives, the gap can be 5-10 percentage points: gross return rate might be 28%, net (refund-only) might be 18%. Net is the right number for CFO P&L modeling because exchanges and store credits don't reduce net revenue. Gross is the right number for ops and 3PL capacity planning because all returns require the same processing work regardless of refund treatment.

how can I reduce returns without hurting conversion?

Six levers in order of impact: (1) Fit content and sizing tools — fit quizzes, True Fit / Fit Analytics integrations, model height/size data on PDPs cut apparel returns 3-6 points. (2) Detailed PDPs with multi-angle photos, video, and customer-submitted try-on photos reduce wrong-product returns 2-4 points across categories. (3) Exchange-first return flows (Loop, Returnly) convert 30-50% of would-be refunds into exchanges, preserving revenue without forcing the customer to repurchase. (4) Return-fee gating for repeat returners (after 3+ returns in 12 months, charge $5-7 per return shipment) without affecting first-time customers. (5) Smaller, more curated assortments — broader catalogs increase choice paralysis and bracketing. (6) Post-purchase comms (sizing confirmation emails, fit tips for the specific item) catch return triggers before the item ships back.