Why is my Shopify return rate so high? A diagnostic
You open your Shopify analytics and see the return rate has crept up another point this quarter. Your gut says something is wrong. You don't know what. Is it the new products? The new photos? The new free-shipping threshold? A new customer cohort? Or is something being exploited?
Before you change your return policy, rewrite product descriptions, or start blocking customers, run this diagnostic. Five steps. At the end you'll know whether you have a fit problem, a product problem, a communication problem, a seasonality problem, or a fraud problem, and which to fix first.
Step 1: compare your rate against your category, not the store average
The single most misleading number in e-commerce is "average return rate." Depending on the source you'll see 10% or 30%. Both are technically true. Neither tells you whether your rate is bad.
A 22% return rate on women's dresses is healthy. A 22% return rate on bedside lamps is a crisis. Return rates vary dramatically by category, and comparing across categories is like comparing inventory turns on perishables to electronics.
Pull your return rate by product type or collection. Not just store-wide. If every category is above its category benchmark, the problem is systemic (policy, shipping, fraud). If one category is blowing out, the problem is narrow (product, fit, or photos in that category).
Output of step 1: one sentence. "My return rate is high/normal in category X, Y, Z."
Step 2: separate legitimate from abusive returns
Most merchants look at one return rate. There are four.
- Sizing returns. Customer wanted it, it didn't fit. Normal for apparel and footwear.
- Product returns. Customer got what they ordered, didn't like it. Signal of a quality or expectation mismatch.
- Damage and not-received returns. Legitimate logistics issues and sometimes INR fraud.
- Abuse returns. Wardrobing, bracketing, serial returners, gift-return fraud. The real cost center.
If you don't already tag returns by reason, start. Every Shopify return carries a reason field. Enforce it through your return portal. Once you have 90 days of categorized data, plot the four rates separately. You'll usually find one or two categories dominating the trend.
Merchants who run this for the first time are consistently surprised by how much of their "high return rate" is actually abuse disguised as normal returns. For deeper analysis, see Return reason clustering.
Step 3: check the recent customer cohort
A return rate doesn't rise because existing customers change their behavior. It rises because your new customer mix is different.
Did you launch on a new acquisition channel in the last 90 days? TikTok, Meta, Google Shopping, Klarna's marketplace, and BNPL-integrated channels all bring customers with measurably higher return rates than organic. BNPL-acquired customers in particular return at 2x to 3x the rate of cash and card customers on apparel.
Did you open a new geography? International customers behave differently. EU and UK customers return at higher rates than US customers on the same SKUs, driven by statutory 14-day cancellation rights. See International return fraud: why your EU customers behave differently.
Did you change your return policy? Any move toward "free returns, no questions asked" will lift your return rate within 2 to 3 weeks. The question is whether conversion lift justifies it. The true cost of a 'no questions asked' return policy walks through the math.
Did you lower your AOV threshold? New customer segments at lower price points often have less price sensitivity per item and return more casually.
A return rate spike without a clear cohort change is the strongest signal that something non-organic is happening. Usually fraud.
Step 4: look for seasonality
Return rates follow predictable curves. Q4 returns spike in January. Wedding dresses return in May. Formalwear spikes after prom. Swim returns after Labor Day.
Pull the last three years of return data and plot the weekly return rate. If your current rate fits the seasonal curve, don't panic. You're seeing a natural pattern. If it's above the curve by 20%+, something has changed.
Also watch the delay distribution. Organic customers return within 7 to 14 days. Wardrobing returns within 2 to 3 days (wore it once, sent it back). Abuse returns cluster at 27 to 30 days, right before the return window closes. A rising share of "day 28 to 30" returns is a fraud signal, not a seasonality signal. See Seasonal return fraud for the full cyclical playbook.
Step 5: find your 1%, the worst offenders
This is the step almost no merchant runs because it feels intrusive. Do it anyway.
Pull a list of customers by 90-day return count. Sort descending. Look at the top 1%. Typically:
- The top 1% of customers by return count drive 30% to 50% of your return losses
- Those customers usually have return-to-order ratios above 60%, meaning they return more than half of what they buy
- Many have multiple accounts registered at the same shipping address with different email addresses (classic multi-account pattern)
If your top 1% looks abusive, that's your answer. The "high return rate" isn't a store-wide problem. It's a small-group exploitation problem. If you don't tag or block those customers, it will continue indefinitely, and they'll recommend your store to others who behave the same way.
For patterns to look for, see Wardrobing: fashion's invisible fraud vector and Emerging return fraud patterns.
Decision tree: what to fix first
Based on the output of steps 1 through 5:
| Signal | Diagnosis | Fix |
|---|---|---|
| High in one category, normal elsewhere | Product or fit problem | Better photos, sizing guide, customer feedback loop |
| High across all categories, no cohort change | Policy too permissive | Tighten return window, add restocking fee on select categories |
| High only in new acquisition channel | Channel mismatch | Filter or pause the channel, adjust messaging |
| High in specific geography | International policy issue | Adjust return policy per region |
| Top 1% driving the lift | Fraud or abuse | Score returns, tag risky customers, block repeat offenders |
| Delays clustering near window close | Organized abuse | Tighten return window, require photos for high-risk categories |
The last two rows are where most merchants stop diagnosing because it feels uncomfortable. It shouldn't. It's just math.
When to bring in scoring
If steps 4 and 5 show the problem is abuse-driven, you have three choices:
- Manually review every return above a threshold. Works until you hit 200 returns a month.
- Block based on hard rules. Catches the obvious cases, misses the sophisticated ones.
- Score every return against multiple signals. Catches both, tunable per shop.
RefundSentry was built for option 3. Every Shopify return is scored against 80+ signals (wardrobing patterns, bracketing, multi-account clusters, refund-method switches, velocity, more) within 2 seconds of the return request. Your team sees the risk zone before they approve the refund. During the private beta, it's free. No credit card, no feature gating.
Most merchants who run this diagnostic find they have a mix: 30% fit issues, 30% product-communication issues, 40% abuse. The first two need product, merchandising, and copywriting fixes. The last 40% needs a scoring system. Don't confuse them. The fixes are different.
Need a structured audit of the last six months of returns? How to audit six months of return fraud without hiring a data team walks through the full methodology.