Ecommerce returns: how return-rate exposure rewrites net revenue and valuation

What "return rate" actually means — and why gross revenue can mislead

Return rate is usually expressed two ways: unit return rate (returned units ÷ shipped units) and dollar return rate (returned order value ÷ gross order value). The two rarely match, because returned items skew toward higher-ticket products and toward categories like apparel where a single order often contains multiple sizes.

A seller's own dashboard almost always reports gross bookings — what was ordered — not what was kept after refunds, cancellations, and chargebacks settle. Treating gross bookings as revenue overstates the top line and, if a buyer doesn't correct for it, overstates every multiple applied downstream.

How big is the exposure, and does it vary by channel

The National Retail Federation's 2025 Retail Returns Landscape, produced with Happy Returns, puts the overall 2025 retail return rate at 15.8% of sales, in line with 16.9% the prior year. Online sales carry a higher rate: an estimated 19.3% of ecommerce orders come back, versus a lower rate for in-store purchases.

That online-versus-in-store gap isn't a one-year blip. NRF's 2023 report with Appriss Retail found the same pattern: a 17.6% online return rate against 10.02% for brick-and-mortar-only sales, with total 2023 returns reaching $743 billion, or roughly $145 million in merchandise returns for every $1 billion in retail sales.

Category matters as much as channel. Shopify's own ecommerce returns data puts the ballpark at close to 20 returns for every 100 products sold online, and singles out apparel and footwear as running well above that average because fit and sizing are hard to judge from a product page — a structural driver, not a policy failure.

Net-of-returns revenue: the reconciliation a buyer actually needs

Before touching a multiple, reconcile the order ledger the same way a diligence process would for any other revenue stream. For each period, document:

• Gross orders shipped, in units and dollars
• Returns, refunds, and cancellations, by reason where available
• Chargebacks and payment disputes
• Restocking fees collected, if any
• Recognized net revenue after all of the above

A platform's "total sales" figure is a gross number by default. Confirm it against the payment processor's net deposits and the accounting system's recognized revenue line — not a marketing dashboard, which may report gross bookings because that number looks better.

Where the real cost hides beyond the refunded dollar

The refund itself is only the first cost. A return also consumes reverse shipping, inspection labor, restocking or liquidation of the unit, and — for a share of returns — outright fraud.

NRF's 2025 data found that 9% of all returns are fraudulent, with retailers reporting rising incidence of overstated return quantities, "empty box" claims, and counterfeit decoy returns.

The 2023 report quantified the drag directly: $13.70 in fraud losses for every $100 of returned merchandise.

None of that fraud loss, restocking labor, or reverse-logistics cost shows up as a separate line on most sellers' P&Ls. It's usually buried inside COGS, shipping expense, or simply netted against gross sales — which is exactly why a buyer has to ask how the return rate was calculated, not just what number the seller quotes.

How buyers adjust SDE for return and refund exposure

Normalizing earnings for return exposure is a two-step process, not a single haircut.

First, confirm the revenue figure used to compute SDE is already net of returns, refunds, and chargebacks. If a seller's stated revenue is gross bookings, SDE built on top of it is inflated before any expense adjustment even happens.

Second, treat the return-rate trend itself as a risk signal. A return rate near the ecommerce baseline that has been flat or improving supports confidence in that net figure. A return rate that is elevated relative to category norms, rising quarter over quarter, or undocumented by reason code, raises the same kind of concern as customer concentration or thin margin — it doesn't necessarily lower the number, but it lowers confidence in the number holding after close.

A one-time spike tied to a specific defective batch or a since-fixed sizing chart is a different risk than a return rate that is structurally high because the product category itself runs hot — apparel being the clearest example. The first is explainable and often already fixed; the second is a durable feature of the business a buyer is acquiring.

Open preview

Category variance a buyer should expect — and why "average" is misleading

A blended 19.3% online return-rate average, per NRF, hides enormous spread. Apparel and footwear sit well above it because of fit uncertainty and multi-size ordering, as Shopify's data notes.

Categories where the product either works or it doesn't — consumables, most electronics accessories, single-SKU commodity goods — typically run below the blended average.

That means comparing a target company's return rate to the 19.3% national online average without first matching category is close to meaningless. The right comparison is category-specific: an apparel store at 22% may be performing better than peers, while a commodity-hardware store at the same 22% is a red flag worth investigating.

Return-policy documentation is diligence evidence, not a footnote

A written return policy and the platform configuration behind it are part of the evidence file, not boilerplate. Shopify's own returns and exchanges documentation describes return and cancellation rules that "control which items are eligible and how long customers have to make a request," alongside separate refund, return-receiving, and exchange actions merchants can configure per order.

For diligence, request:

• The written return/refund policy as published to customers
• The platform's configured return window and any restocking-fee rules
• Whether self-serve automated returns are enabled, and since when
• A return-rate export by reason code and by SKU or category, not just a headline percentage
• How refunds are timed relative to the return being received versus initiated

A policy that looks generous on paper (free, unlimited returns) but has no corresponding return-rate discipline in the data is itself a signal — either the number is being managed elsewhere, or the exposure hasn't been priced into margin at all.

How RSW treats return exposure in a valuation

RSW starts from normalized, net-of-returns earnings, the same way it normalizes for inventory and COGS risk elsewhere in an ecommerce valuation. It does not apply a blanket percentage haircut for a given return rate, because the right benchmark depends on category, and a rate close to category norms with clean documentation is simply a normal cost of doing business.

What changes confidence is verification: whether reported revenue was already net of returns, whether the return-rate trend is stable or deteriorating, and whether the seller can produce category-matched benchmarks rather than a single self-reported number. Elevated, undocumented, or rapidly rising return exposure increases uncertainty the same way an unreconciled channel-revenue map does — not because returns are inherently bad, but because unverified numbers are.

Buyer checklist

Request, at minimum:

• Gross orders and net (post-return) revenue, reconciled by period
• Return rate by unit and by dollar, with the calculation method disclosed
• Return reason codes, or an explanation for why none are tracked
• Fraud/abuse rate among returns, if tracked
• Written return policy plus the live platform configuration behind it
• Category benchmark comparison, not just the store's own trailing average

FAQs

What is the average ecommerce return rate?

NRF and Happy Returns put the 2025 online return rate at an estimated 19.3%, against a blended 15.8% across all retail. Category matters far more than the blended average — apparel runs well above it, commodity and single-SKU goods typically run below it.

How do I calculate an ecommerce return rate?

Divide returned units by shipped units for a unit rate, or returned order dollars by gross order dollars for a dollar rate. The two figures diverge when higher-ticket items or multi-item orders return disproportionately, so a buyer should ask which method a seller used before comparing it to a benchmark.

Does a high return rate always lower a business's value?

Not automatically. A return rate near category norms with clean, documented reconciliation is a normal cost of doing business. What raises risk is an unverified figure, gross revenue reported as if it were net, or a rate that's rising without an identified cause.

Should ecommerce stores offer free returns?

That's a merchandising and conversion decision outside RSW's valuation scope. For diligence purposes, what matters is whether the policy's cost is actually showing up in the store's margin and return-rate data, not whether free returns exist in principle.

Is there seasonality to return rates?

Yes — NRF's data shows holiday-period returns running higher than the annual average, around 17% of holiday sales in 2025. A trailing-twelve-month return rate smooths that seasonality out; a rate calculated from a single quarter may not.

Continue through the RSW silos

Start with the ecommerce valuation pillar, then compare Shopify store valuation and inventory and COGS risk, which covers the unsold-inventory side of the same margin question returns sit on.

Repeat purchase rate and supplier concentration round out the durability picture, and selling through TikTok Shop applies the same net-revenue discipline to a newer channel.

Get a range, then verify what's net of returns

Use the Real Site Worth website value calculator for an automated starting range, or go straight to the ecommerce store valuation tool built for owner-input revenue and margin. Before any transaction, reconcile gross orders to net revenue, pull the return-rate trend by category, and confirm the policy on paper matches the cost showing up in the numbers.