Your return rate cannot answer the one question that matters: which customers are worth keeping once returns are priced in. Two short chapters on why, a tool that runs on your own numbers, and the move.
Retail’s most-cited fraud figure was published two ways in the same year, from bases $205 billion apart, and its own publisher says the years are not comparable. The only returns number you can trust is your own.
3 min read · October 26, 2026
NRF and Happy Returns$890B total returns
× ~11.6% fraud rate
Appriss and Deloitte$685B total returns
× 15.1% fraud rate
Same year, $205B apart
Both land on $103B
Monday morning
Most reporting treats returns as an operations line. Priced customer by customer, they redraw who your best customers actually are.
2 min read · October 26, 2026
$560.00 gross
+$227.00
Customer A
1 of 8 units returned
$560.00 gross
−$62.00
Customer B
9 of 14 units returned
Monday morning
See who is driving your returns, free, by tomorrow morning.
Price one return at your own costs, then score a customer on net contribution after returns. It runs here, on your own numbers. Nothing to download, nothing to sign up for.
The refund is the small part. Fill in your own figures; the worked example is an $80 apparel return at 60% gross margin.
Original price minus what the returned unit resells for.
Loaded cost of one return
$31.80
On top of handing the refund back.
What the order swings by
−$79.80
It earned $48.00 when it shipped. The return reverses that and costs $31.80 more.
Kept orders to earn it back
1.7
Clean, kept orders at the same margin to pay for one return.
If you cannot itemize, the all-in cost of an apparel return is commonly put at 20% to 40% of the item’s price. Use 30% as a first estimate.
Pick a customer, or a segment, and run their lifetime numbers. Keep rate is the bracketing signal your return rate hides.
Price minus product cost, per unit.
From the first calculator.
Net contribution after returns
−$42.20
$120.00 earned, $127.20 of returns, $35.00 to acquire.
Keep rate
36%
Units kept over units ordered. Below about 75% is bracketing.
Profile
Serial bracketer
Add friction: a return fee, final-sale nudges, size guidance.
Among merchants that started charging for returns, 47% saw complaints rise, 37% lost customers and 34% saw order value fall. Blanket fees tax the loyal high-returners you most need to keep. Aim the lever instead.
NRF and Happy Returns, 2025 Retail Returns Landscape.
Protect. Free, frictionless returns.
A generous return experience is what makes them loyal.
Add friction: a fee, final-sale nudges, size guidance.
Recovers margin from the cohort driving reverse-logistics cost.
Grow share of wallet. Cross-sell, replenish.
Cheapest customers to serve, most headroom.
Stop re-acquiring on paid.
You lose money each time you buy them back.
The arithmetic above is the arithmetic Tacet runs on your whole customer base, every night, from your own refunds.
Every refund priced into the customer who made it.
From your own Shopify refunds, every night, with your own handling cost per return and the margin on what came back. No industry figure in the way.
The few behind half of the refunded dollars, by name.
Who is driving your returns, with Forward CLV beside each, so the loyal high-returner and the costly one stop looking alike. A return band on every Klaviyo profile.
Playbooks read in profit after refunds.
Every playbook is measured against a held-back group after discounts and refunds, so a send that reached the bracketers cannot look like it worked.