Skip to main content
Card class: Non-HeroCategory: Ecommerce Platform

At a glance

Percentage of BigCommerce orders cancelled before fulfilment over 30 days. The leading indicator of operational and customer-experience problems, cancellations happen before the customer receives the product, signalling either inventory issues, fraud blocks, or buyer’s-remorse from sub-optimal post-order communication.
What it countsCOUNT(orders WHERE status = 'Cancelled') ÷ COUNT(all orders) over the period. Calculated as a percentage of orders created in the window. Some configurations also count Declined status; this card uses Cancelled only by default.
VAT / tax treatmentn/a (rate metric).
Shippingn/a.
Discountsn/a.
RefundsCancellations and refunds are distinct on BC. Cancellation = order voided before fulfilment; refund = order fulfilled then customer returned. Cross-reference Refund Rate.
Cancelled / voided ordersThis card is the cancellation view.
Currencyn/a (currency-agnostic).
Channels / sourcesAll channels contribute. Marketplaces often have higher cancellation rates (“buyer changed mind”, “unpaid item”); web cancellations are typically merchant-initiated (fraud, OOS, address issue).
Cancellation reasonsBC tracks reasons via the order log: customer request, payment failure, OOS, fraud detection, duplicate order, other. Pair with reason-breakdown view for diagnosis.
B2B Edition noteB2B portal cancellation rates are typically <1% because orders are pre-approved by account managers. Sudden B2B lifts may indicate a payment-terms or portal usability problem.
Healthy benchmarks<2% (excellent), 2-4% (acceptable), 4-7% (needs attention), >7% (problem). Marketplaces typically run 5-15% due to “buyer changed mind”; web should run lower.
Time window30D vsP (30D vs prior 30D for trend)
Alert trigger>3%
Sentiment keycancellation_rate
Rolesowner, operations

Calculation

COUNT(status=Cancelled) / COUNT(_id)
  WHERE date BETWEEN [period_start, period_end]

Worked example

A US homewares brand on BigCommerce Pro, 30-day window 14 Apr 26 to 14 May 26.
ChannelOrders createdOrders cancelledCancellation rateTop reasons
Stencil web (channel_id = 1)4,820962.0%Fraud-rule rejection (38), customer request (28), duplicate order (12), OOS at fulfilment (10), other (8)
Amazon Channel Manager1,14014212.5%Buyer cancelled (84), payment timeout (32), seller cancelled (16), other (10)
eBay Channel Manager4807816.3%Unpaid item (52), buyer cancelled (18), seller cancelled (8)
POS (terminal A + B)28041.4%Customer changed mind in-store (3), card-decline retried successfully (1)
B2B Portal9200%All B2B orders fulfilled
Store-wide6,8123204.7%Mixed; marketplace-driven
What’s interesting:
  1. Web at 2% is healthy benchmark territory. Within that 96 cancellations: 38 are fraud-rule rejections (good, that’s the system working), 28 customer requests (manageable), 22 operational issues (duplicate, OOS, other). The actionable subset is the operational 22; investigate root causes.
  2. Amazon at 12.5% is normal-high for marketplace. Amazon’s “buyer cancelled” workflow allows easy cancellation up to ship; many buyers cancel after better-deal-found. The 32 payment-timeout cancellations are more concerning, indicates Channel Manager’s Amazon-payment-confirmation flow has friction. Investigate.
  3. eBay at 16.3% is on the high end. “Unpaid item” cancellations (52) are eBay’s auction / buy-it-now mechanic where buyers commit but don’t pay. Action: enable immediate-payment requirement on eBay listings; this typically halves the unpaid-item rate.
  4. POS at 1.4% is excellent. In-store customers who reach the till usually complete; the 4 cancellations are mostly card-decline retries that succeed on second attempt.
  5. B2B at 0% is the gold standard. Pre-approved orders from established accounts essentially never cancel. If B2B cancellations rise above 1%, there’s an account-relationship or portal-usability problem.
  6. Store-wide 4.7% is dragged up by marketplace channels. Web alone is 2%; marketplace alone is 13.4%. Compare like-for-like, web cancellation should be evaluated against web benchmarks (<2.5%), marketplace against marketplace benchmarks (<15%).
Action priority order:
  1. Web cancellation root-cause audit identify the 22 operational cancellations (duplicate, OOS, other) and fix systematically.
  2. eBay immediate-payment enforcement flip the listing setting; cuts unpaid-item rate in half typically.
  3. Amazon payment-timeout investigation Channel Manager’s Amazon-payment flow needs review.
  4. Set per-channel cancellation alerts for any channel exceeding its benchmark by 50%.
  5. Quarterly: review fraud-rule effectiveness 38 web cancellations from fraud-rule are healthy if they’re catching real fraud; if they’re false-positives, revise rules to recover the lost orders.

Sibling cards merchants should reference together

CardWhy pair it with Cancellation Rate
Cancelled Over TimeThe trend view; this card is the rate, that card is the cancellation count over time.
Refund RateCancellation = pre-fulfilment, refund = post-fulfilment. Both are revenue-loss signals; pair for the full leakage picture.
BC Decline RatePayment-decline rate; correlates with payment-failure cancellations.
BC Incomplete RateIncomplete (cart abandoned) is upstream of cancellations; high incomplete + low cancellation = abandonment problem; high cancellation + healthy incomplete = post-order problem.
Total RevenueThe denominator context; 5% cancellation on 100krevenueis100k revenue is 5k of lost gross.
BC Channel Refund RatePer-channel refund pattern; combined with this card gives full per-channel revenue-leakage view.
BC Alert Refund Rate SpikeThe movement alert; configure analogous cancellation-rate-spike alert.
shopify.cancellation_rateCross-platform reference.

Reconciling against the vendor’s own dashboard

Where to look in BigCommerce Control Panel: Orders → All orders filter by status Cancelled or Declined, and date range. The list count divided by total orders gives the rate manually. Analytics → Sales on Pro / Enterprise plans shows order counts by status; the cancellation count is implied but not as a percentage. Why our number may legitimately differ from BC:
ReasonDirection
Cancelled vs Declined. We use Cancelled only by default; some BC reports include Declined. Configure to match.Configurable
Time-zone. BC uses store time zone; we use UTC.Boundary-day differences
Cancellations vs voids. Some payment processors report “voided” while BC marks as Cancelled; the underlying state is the same.None at BC level
Channel coverage. We include all channels by default; BC may filter.Different totals
Multi-currency. Rate is currency-agnostic; no FX impact.None
Cross-connector reconciliation (when payment processors connected):
CardExpected relationshipWhat causes legitimate divergence
stripe.stripe_decline_rateStripe declines correlate with payment-failure cancellations.Stripe’s view is gateway-side; this card is post-order-creation.
paypal.pp_cancelled_ratePayPal cancellations should correlate with web-channel cancellation.PayPal sees only PayPal-paid orders.
amazon_sp.amazon_cancellation_rateAmazon Seller Central cancellation rate; should match Amazon channel slice.Different time-zone and refund-vs-cancellation handling.
The cancellation-rate metric is BC-aligned with similar metrics on Shopify and Adobe Commerce; the field shape (BC’s Cancelled status) is platform-specific but semantics equivalent.

Known limitations / merchant FAQs

My web cancellation rate is 8%, what’s wrong? Likely fraud-rule miscalibration or systemic OOS-at-checkout. Run the reason breakdown: (1) if fraud-rejection is >40% of cancellations, your fraud rules are too tight, recover real customers; (2) if OOS-at-fulfilment is >25%, your inventory sync between BC and warehouse has lag (orders accept then can’t fulfil); (3) if customer-request is >40%, post-order experience is failing (no shipping confirmation, slow ship, etc.). How is cancellation different from refund? Cancellation = order voided before fulfilment (no money exchanged for goods, customer never received product). Refund = order fulfilled, customer received product, then returned for refund. Both are revenue-leakage but root causes are completely different. Cancellation root causes are pre-purchase and pre-fulfilment; refund root causes are product-quality, expectation-mismatch, or fulfilment-damage. Why is my marketplace cancellation rate so high? Marketplaces have buyer-friendly cancellation workflows. Amazon allows cancellation up to ship; eBay’s “unpaid item” mechanism cancels uncomplete-payment orders automatically. Marketplace cancellation rates of 8-15% are normal; below 5% suggests strict listing settings (good for cancellation rate, may suppress sales). Should I count “Declined” status as cancellation? Configurable. Some merchants count both as cancellation; we default to Cancelled only. Declined orders are typically payment-rejection (fraud-rule, card-decline); they’re a slightly different cohort from cancelled orders. If your fraud rules are aggressive, including Declined inflates the rate; excluding gives a cleaner customer-experience signal. My cancellation rate dropped sharply, possible cause? Common: (1) fraud rules loosened (catching less, allowing more orders to complete); (2) OOS-at-checkout fixed (better inventory sync); (3) marketplace-side mechanic change (Amazon tightened cancellation window); (4) ad-traffic mix shift toward higher-intent customers. Verify by reason breakdown. Should I auto-cancel duplicate orders? Yes for confirmed duplicates (same email, same items, same shipping, within 30 minutes). The customer almost always intended one order; the duplicate cancellation is a courtesy. Don’t auto-cancel orders from the same customer hours apart, those may be intentional. My fraud-rule rejections are high, what’s the trade-off? Tightening fraud rules reduces fraud-loss but also reduces legitimate-customer orders. The trade-off: 1% fraud rate (loss) vs 5% extra cancellation rate (lost revenue). For most stores, accepting some fraud loss to recover real-customer orders is net-positive. Audit the fraud-rejected orders monthly: how many were genuinely fraudulent vs false-positives? My B2B cancellation rate climbed from 0% to 3%, why? Investigate immediately. B2B cancellations are unusual; common causes: (1) credit-line / payment-terms issue (account hit credit limit, order auto-cancelled); (2) portal usability problem (B2B customers misordering then cancelling); (3) account-manager change (orders being approved that shouldn’t have been). Time-sensitive issue. Should I send a “we noticed you cancelled” email? Yes for customer-request cancellations, no for fraud / OOS / system cancellations. The customer-request email asks “what went wrong?” and offers a small incentive (5% off) to retry. Conversion of these emails is typically 8-15%, recovering meaningful revenue. Don’t send for fraud-rejected orders; that’s actively counter-productive. My “buyer changed mind” cancellations are climbing on Amazon, what changed? Could be: (1) Amazon listing showing wrong size / colour (drives buyer regret); (2) competitor lower price appearing during the buyer’s hesitation window; (3) shipping-time lengthened (Amazon-prime expectations). Audit the listings; check pricing competitiveness; verify Channel Manager fulfilment SLA. Multi-currency: cancellations vary by currency? Yes typically. USD orders may have different fraud-rule sensitivity than EUR. Filter by currency for region-specific cancellation patterns; especially relevant for stores serving multiple regions. Why doesn’t this card have a built-in alert? The default >3% alert exists in the manifest. The threshold is configurable; some merchants set 5% (more lenient for marketplace-heavy stores), others 2% (stricter for web-only stores). Adjust under Vortex Mind → Settings → Alerts.

Tracked live in Vortex IQ Nerve Centre

Cancellation Rate is one of hundreds of KPI pulses Vortex IQ tracks across BigCommerce and 70+ other ecommerce connectors. Nerve Centre runs the detection layer; Vortex Mind investigates the cause when something moves; Ask Viq lets you interrogate any number in plain English. Start for free or book a demo to see this metric running on your own data.