Returns are costing you $120 per order. Here's the math.

Most DTC brands budget $12-15 per return. That's shipping. The actual cost is 8x higher. ## The full cost of a single return | Cost line | $50 AOV | $75 AOV | $150 AOV | |-----------|---------|---------|----------| | Refund amount | $50 | $75 | $150 | | Lost margin (60% GM) | $30 | $45 | $90 | | Return shipping | $10 | $10 | $12 | | Processing labor (30-45 min @ $25/hr) | $15 | $15 | $15 | | Customer service (2.3 avg touchpoints @ $4) | $9 | $9 | $9 | | Chargeback risk (amortized) | $6 | $6 | $6 | | **Total** | **$120** | **$160** | **$282** | At a 20% return rate on 10,000 orders/month, that's 2,000 returns. At $120 average: **$240,000/month. $2.88 million/year.** ## The costs most brands miss **Repeat purchase impact.** A customer with a bad return experience is 3x less likely to reorder. If your average customer makes 2.5 purchases and you lose the second purchase on 30% of returners, that's $37.50 in lost future revenue per return (at $75 AOV). **Review damage.** 15% of customers who have a slow return experience leave a negative review. Each negative review costs an estimated $50-100 in lost conversions. **Restocking waste.** Only 50-70% of returned items can be resold at full price. The rest go to liquidation (30-50 cents on the dollar) or waste. Add these and the true all-in cost is closer to $150-180 per return on a $75 AOV product. ## Where the money goes The refund and margin loss are fixed. You can't change those. The controllable cost is the $35-40 in processing, support, and chargeback risk. That's labor and friction. | Controllable cost | Current | With procedure | |-------------------|---------|----------------| | Processing labor | $15 | $1-2 | | Customer service touches | $9 | $2-3 | | Chargeback risk | $6 | $0-1 | | **Total controllable** | **$30** | **$4-6** | At 2,000 returns/month, cutting $25 per return saves **$50,000/month. $600,000/year.** ## The triage model Not every return needs the same review. **Auto-approve (60-70% of returns).** Order within return window, refund under $50, item not final sale, customer has fewer than 3 returns in 6 months, reason is defective/damaged. Process instantly. No human touches it. **Quick review (20-25%).** One policy condition is borderline — day 31 of a 30-day window, or customer has 3 prior returns. All context pre-gathered, human approves or overrides in 10 seconds. **Full investigation (5-10%).** High-value item, repeat returner, suspected abuse. Human reviews the full picture. 60-70% of returns don't need a human. Processing them manually is paying $15 in labor for a decision a machine can make in seconds. ## What to measure | Metric | Target | Red flag | |--------|--------|----------| | Cost per return | Under $80 | Over $120 | | Time to refund | Under 24 hours | Over 72 hours | | Auto-approval rate | Over 50% | Under 30% | | Return-to-chargeback rate | Under 2% | Over 5% | ## The return rate myth The right goal isn't reducing your return rate. It's reducing your cost per return. A 20% return rate at $40 per return is better than a 15% return rate at $120 per return. The first business processes returns efficiently and retains customers. The second makes returns painful and loses them. --- *DeepMerge automates return triage for DTC brands. AI agents evaluate returns against your policy in seconds — auto-approving the clear cases, escalating the edge cases, and processing fast enough that chargebacks don't happen.*