At a glance
The percentage of paid BigCommerce orders that ended atstatus = RefundedorPartially Refunded. Headline post-purchase regret signal: how often does a paid order get returned for money back? Industry baseline: apparel 20-30%, homewares 5-12%, consumables 2-5%; above-category average suggests product-quality, expectation-mismatch, or fulfilment issues.
| What it counts | COUNT(orders WHERE status IN ('Refunded', 'Partially Refunded')) ÷ COUNT(paid orders in period) × 100. The denominator excludes Cancelled, Incomplete, Declined (no payment captured). |
| VAT / tax treatment | The dollar version uses refundedAmount (BC’s standard refund accounting field, tax-inclusive). |
| Shipping | Shipping cost is included in total_inc_tax; whether it’s refunded depends on the merchant’s policy (US: usually no; UK: usually yes per consumer law). |
| Discounts | Discounts already deducted from the original total; refunds reverse the post-discount amount. |
| Refunds | This is the refund metric; the rate is the count ratio. |
| Cancelled orders | Excluded from denominator (different lifecycle). |
Incomplete / Declined orders | Excluded from denominator (no payment captured). |
| Currency | n/a (rate metric). |
| Channels / sources | All channels combined. Marketplace channels often have higher refund rates due to lower-friction return policies (Amazon’s A-to-Z guarantee). Per-channel breakdown lives at BC Channel Refund Rate. |
Partially Refunded semantics | Counted as 50% credit toward refund rate. A partial refund is still a customer-experience event but represents lower revenue impact than a full refund. |
| Time window | 30D vsP (default 30D vs the prior 30D) |
| Alert trigger | >5% (or +25% vsP), dual-trigger: absolute level above 5% OR a 25% relative jump from prior period. |
| Sentiment key | refund_rate |
| Roles | owner, operations |
Calculation
Worked example
A US homewares brand on BigCommerce Enterprise. The 30-day window covers 14 Mar 26 to 12 Apr 26.| Cohort | Order count | Refund count | Refund rate |
|---|---|---|---|
| All paid orders | 2,948 | 198 | 6.7% |
Web (channel_id = 1) | 2,212 | 122 | 5.5% |
| Amazon Channel Manager | 612 | 64 | 10.5% |
| Facebook Shop | 88 | 8 | 9.1% |
| POS | 36 | 4 | 11.1% |
| Prior 30D rate | 2,840 | 168 | 5.9% |
- Headline 6.7% is above the homewares baseline (5-12%) but elevated, the alert fires because the absolute level exceeds 5% AND the period-over-period delta (+13.6%) is approaching the 25% threshold.
- Amazon’s 10.5% is the killer. Amazon’s A-to-Z guarantee makes refund-claim friction near-zero for buyers; Amazon-channel refund rates run 1.5-2x web rates structurally. The store’s combined rate is dragged up by Amazon’s higher rate.
- Web refund rate of 5.5% is healthy. If the merchant could de-risk Amazon (more careful product descriptions, more accurate sizing, additional product images) the headline rate would drop to ~6.0% even with current volumes.
- The vsP delta (+0.8 pp absolute, +13.6% relative) is the leading indicator. Whatever caused the rise in this period continues unless addressed. Common causes: a new SKU launched with quality issues; a fulfilment-speed problem leading to “arrived too late, refund please”; a product description edit that misled customers.
- Run BC Refunded Products to identify which SKUs are driving the rise. Almost always, 3-5 SKUs concentrate the refunds.
- Audit Amazon listings for those high-refund SKUs, expectation mismatches drive most Amazon refunds.
- Cross-check fulfilment speed, refunds correlate with delays.
- Review recent product descriptions and images on web to see if any changed near the start of the period.
Sibling cards merchants should reference together
| Card | Why pair it with Refund Rate |
|---|---|
| Refund Value | The dollar twin; rate × AOV ≈ refund value. |
| BC Refunded Products | SKU-level breakdown; identifies which products drive the rate. |
| Refund Count | The numerator. |
| BC Channel Refund Rate | Per-channel breakdown. |
| Cancellation Rate | The pre-fulfilment cancellation companion. |
| Fulfillment Rate | Slow fulfilment drives refunds; co-track. |
stripe.stripe_refund_rate | Stripe-side refund rate; should align with web channel slice. |
paypal.pp_refund_rate | PayPal-side equivalent. |
Reconciling against the vendor’s own dashboard
Where to look in BigCommerce Control Panel: Orders → All orders, filter by statusRefunded and Partially Refunded for the period; divide by total non-failed orders. Should match this card to within ±0.5 pp. Analytics → Insights on Plus / Pro plans has a Returns dashboard with similar data.
Why our number may legitimately differ from BC Control Panel:
| Reason | Direction |
|---|---|
Partially Refunded weighting. We use 50% credit; BC reports may use 0% or 100%. | Variable |
| Time zone. UTC vs store-local. | ±0.5 pp |
| Refund timing. Refunds happen days/weeks after the original order; we attribute the refund to the refund date, BC may attribute to the order date. | Different period selection results |
| Test refunds. Not currently filtered. | Marginal |
bigcommerce.refund_rate = refund_count ÷ order_count × 100
Component cards (these should be self-consistent, if they’re not, it’s a sampling or rounding issue, not real divergence):
Known limitations / merchant FAQs
Why is my Amazon refund rate so much higher than web? Amazon’s A-to-Z guarantee makes refund-claim friction near-zero; buyers can request refunds with a click. Web checkout typically requires an email or phone call, so customers self-select to “actually I’ll keep it” before refunding. Industry pattern: Amazon refund rates run 1.5-2x web rates structurally. Don’t try to match web; manage Amazon refunds via better product-page expectations (richer images, accurate sizing, full descriptions). My rate spiked, how do I diagnose? Run BC Refunded Products. 80% of the time, 3-5 SKUs concentrate the refunds. Look at those SKUs: did the description change? Is one specific size / colour failing? Did a fulfilment delay land in this period? Cross-reference with Fulfillment Rate for the speed dimension. Why is my homewares refund rate higher than apparel benchmarks I see online? Most “ecommerce average refund rate” benchmarks are apparel-weighted; apparel runs 20-30% by category nature (size, fit, colour mismatch). Homewares should be 5-12%; consumables 2-5%. Use category-specific benchmarks rather than ecommerce-wide averages. Should I tighten my returns policy to reduce the rate? Generally no. Tighter policies (shorter window, restocking fees, customer-pays-return-shipping) reduce refund rate but also reduce conversion (customers shop elsewhere with better policies). Net revenue typically falls. Fix the cause of refunds (product-page expectations, fulfilment quality) rather than the symptom (refund-policy friction). My refund rate dropped, is that good? Usually yes, but check whether (a) order volume also dropped (rate looks better simply because fewer orders to refund), or (b) you tightened policy and now customers are keeping unwanted items but reviewing badly. Cross-check Total Revenue and customer-review trends to confirm the drop is real improvement. Why excludeCancelled and Incomplete from the denominator?
Both populations had a different lifecycle, neither completed payment in the same way as a captured order, so including them dilutes the rate. The card focuses on “of orders that successfully captured cash, how many were later refunded?”, which is the operational question.
Multi-currency stores: does the rate look different by currency?
Yes; UK / EU customers refund more often than US customers due to consumer-protection law differences. EU consumers have a 14-day no-fault return right (the “cooling-off period”) that doesn’t apply in the US. EU/UK refund rates run 1-3 percentage points higher structurally.
Should this card include Disputed chargebacks?
Currently no; chargebacks have a different lifecycle and operational response. The Vortex Mind dispute-rate card covers chargebacks separately.
Why doesn’t BC have a single ‘Refund Reason’ breakdown?
BC doesn’t natively capture refund reasons in a structured field; reasons are typically free-text in staff_notes or in customer-service email threads. The Vortex Mind refund-cause-analysis report uses ML to categorise free-text reasons into structured categories.
A 1-percentage-point drop saved how much?
For a 50k of recovered revenue per year (refunded value × rate reduction). Worth fighting for. Refund-rate improvements compound: better product pages reduce refunds AND improve conversion AND reduce customer-service load, all from the same fix.