Returns: The $890 Billion Problem Nobody Wants to Own

Somewhere between one in six and one in three of everything e-commerce sells comes back. US retail returns run roughly 16.9% of total sales — about $890 billion — and online apparel pushes 25–35%. Yet returns remain the orphan of DTC org charts: marketing owns acquisition, ops owns outbound, finance owns the write-off, and the full reverse-logistics P&L belongs to no one. That orphan status is exactly why it's one of the last large, unworked margin levers in e-commerce — and one of the first places we look in diligence.

The true cost of a $60 return

Cost Component	Typical
Return shipping (label)	$6–9
Inspection & processing labor	$3–5
Repackaging / refurb	$1–4
Markdown or liquidation loss	$5–20
Payment fees (non-refundable)	$1.80
Total margin destruction	$17–40 on a $60 item

Now run the portfolio math: a brand doing $5M with a 22% return rate is processing $1.1M of reverse flow; at a blended 30% cost-of-return, that's $330K of margin — frequently more than the brand's entire net profit — flowing through a process nobody owns. This is why two brands with identical toplines and "similar" return rates can differ by hundreds of thousands in true earnings, and why we rebuild returns economics from raw data in every diligence rather than accepting the netted revenue line.

The prevention third

Industry analyses consistently find a quarter to a third of returns are avoidable — fit and expectation failures, not product defects. The playbook is unglamorous and works: merciless size-guide accuracy (fit-predictor tools cut apparel returns double digits), photography and copy that show true color/scale/texture, review mining for recurring "expected bigger/smaller" signals fed back into PDPs, and packaging that survives the journey. Each point of return rate on a $5M apparel brand is worth roughly $15–25K of recovered margin annually — prevention compounds like retention does.

Recovery, fraud, and the policy frontier

For the returns that happen anyway: disposition speed is everything (every week in reverse limbo is markdown decay); returnless refunds are rational below a cost threshold and in low-resale categories; resale/open-box channels recover 40–70% where liquidators recover 10–20%. On fraud — a multi-billion-dollar and growing problem — the modern stance is tiered trust: generous policies for high-LTV honest cohorts, friction (photo verification, label tracking, refund-after-inspection) reserved for statistical outliers. Blanket generosity and blanket suspicion both lose money; segmentation wins. Policy itself is a pricing decision: free returns lift conversion and lift return rates, and the right answer is category math, not ideology.

Returns are where gross margin goes to die quietly. The operators who give the corpse an owner get the margin back.

Sources: NRF/Appriss returns surveys (16.9% / ~$890B); category return-rate benchmarks (apparel 25–35%); reverse-logistics cost studies; returns-fraud industry reporting. Companion: State of DTC 2026.