At a glance
The percentage of orders in the rolling window that ended in a “transaction defect” by eBay’s definition. The single threshold most likely to flip a seller out of Top Rated standing: TRS requires defect rate ≤0.5%. Above 0.5% and you’re out of TRS, lose the 10% FVF discount, lose the search lift, lose the badge.
| What it counts | COUNT(transactions WITH at least one defect) ÷ COUNT(total transactions) × 100. Defects include: cases closed without seller resolution, sellers cancelling orders out-of-stock, late shipment with no tracking, INR cases auto-decided against the seller, “Item Not as Described” returns the seller refused. |
| Listing-format scope | All formats. Auctions, fixed-price, and Best-Offer-resolved orders are equally eligible to receive a defect. |
| GMV / fees framing | Not applicable, this is a count ratio. Each defect-bearing order counts as 1 regardless of order value. A £5 order with a defect counts the same as a £500 order with one. |
| Promoted Listings | Promoted-driven orders count toward both numerator and denominator identically. Scaling Promoted volume on defect-prone SKUs can mechanically push the rate up. |
| Multi-site aggregation | Computed per marketplaceId. The card displays each connected site separately; there is no “global” defect rate (eBay evaluates seller-standing per-site). |
| Currency | Not applicable, count ratio. |
| Best-Offer-resolved orders | Counted in numerator and denominator identically to BIN. |
| Refunds | A seller-issued refund before an INR or “Item Not as Described” case escalates can prevent the defect being recorded. Refunds after the case auto-resolves do not retroactively reverse the defect. |
| Cancellations | Seller-cancelled / out-of-stock cancellations register as a defect against the cancelled transaction. Buyer-cancelled-before-payment cancellations are excluded from both numerator and denominator. |
| Time window | 30D vsP (display window). eBay’s official defect rate is computed over rolling 12 months for established sellers, 3 months for newer sellers; both are exposed via the API and shown alongside the 30D figure. |
| Alert trigger | >0.5% (the Top Rated cap), driven by sentiment_key: feedback_score. The 0.5% line is hard, eBay does not grade. |
| Roles | owner, operations |
Calculation
Calculated automatically from your eBay data. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.Worked example
A UK collectibles seller running ebay.co.uk only. Period: 02 Apr 26 to 01 May 26 (30D). Rolling 12-month window also reported by eBay’s API.| Defect type | Count (30D) | Count (12M rolling) | TRS-eligible? |
|---|---|---|---|
| Cases closed without seller resolution | 2 | 18 | No (over 0.3% cap by itself) |
| Seller-cancelled out-of-stock | 4 | 31 | n/a, separate cap |
| Late shipment with no tracking | 1 | 8 | n/a, separate cap |
| INR auto-decided against seller | 3 | 22 | n/a, contributes to cases-without-resolution |
| Defects (de-duplicated) | 9 unique transactions | 74 unique transactions | |
| Total transactions | 1,800 | 18,400 | |
| Defect rate | 9 / 1,800 = 0.50% | 74 / 18,400 = 0.40% | At cap on 30D, below cap on 12M |
- The 30D rate is on the cliff edge. A single additional defect in the next 24 hours pushes 30D defect rate to 0.56%, breaching the 0.5% cap. The Vortex IQ Nerve Centre is engineered to fire the alert at 0.5%, before the seller silently steps over.
- The 12M rate is what eBay actually evaluates against. TRS standing is checked monthly on the 20th using the 12-month window for established sellers. The 30D is a leading indicator. The seller is currently safe at 0.40% over 12M but trending upward; the 30D rate is a 0.10-point premium over 12M, signalling the trend is worsening.
- De-duplication matters. The four defect categories sum to 10, but only 9 unique transactions; one transaction generated both an INR auto-decision and a “Cases closed without seller resolution” defect (they’re the same incident counted twice in the row counts). eBay’s API returns the de-duplicated count.
- The four sub-caps run separately. Even at 0.40% overall defect rate, breaching the 0.3% cap on cases-without-seller-resolution alone (currently at 18 / 18,400 = 0.098%, safe) would lose TRS. The headline rate is one of four guard-rails; pair with Cases without Resolution, Late Shipment Rate, and Seller Standards for a full read.
- Recovery requires the rate to drop, not just the count. With the rolling window, defects persist for 12 months in the calculation. A clean 30 days won’t recover TRS; a clean 12 months will. Sellers who lose TRS in May typically don’t recover it until the offending transactions roll out of the window the following May.
Sibling cards merchants should reference together
| Card | Why pair it with Defect Rate |
|---|---|
| Cases without Resolution | Separate sub-cap (0.3%). Hits TRS standing independently of overall defect rate. The hardest-to-recover defect category. |
| Open INR Cases | The leading indicator. Each unresolved INR over 8 days auto-converts into a defect. Resolve cases here to prevent defect-rate movement. |
| Late Shipment Rate | Separate 3% TRS cap. Distinct mechanism but same fail consequence. |
| Seller Standards | The cliff outcome. 0.5% defect-rate breach equals automatic Above-Standard demotion. |
| Health Score | The rolled-up composite. Defect rate is one of four inputs (25% weight). |
| Marketplace Fees | The economic tail of a TRS loss, watch fees step up by ~10% the day after a TRS demotion. |
| Total Revenue | Economic context for defect-rate impact. A 0.7% defect rate on £200k of monthly GMV is more urgent than the same rate on £20k. |
| Amazon Account Health | Marketplace peer. Amazon’s defect-equivalent is the “Order Defect Rate” (ODR), capped at 1%. Different threshold, similar mechanism. |
Reconciling against the vendor’s own dashboard
Where to look in eBay Seller Hub:Seller Hub → Performance → Seller Level The “Transaction defect rate” tile shows the 12-month rolling figure that eBay uses for its monthly evaluation.For per-defect drilldown: Performance → Service Metrics shows the rates of “Item Not as Described” and “Item Not Received” claims, the components feeding defect rate. Timing, settlement, and reporting-lag table:
| Topic | Detail |
|---|---|
| Timezone | Defect timestamps are recorded in UTC; Seller Hub displays in account timezone (Pacific Time for US accounts). The 12-month window is end-of-day account-timezone for boundary handling. |
| Settlement / payout impact | A high defect rate causing a TRS-to-Above-Standard demotion shifts payouts from near-instant to a 1 to 5 day funds hold per order. Demotion to Below Standard introduces a 30-day funds hold. |
| Promoted Listings cost reporting lag | None, ad costs aren’t relevant. |
| API throttling | The Sell Analytics API returns the rate every 15 minutes; webhook events from new defect occurrences push within 30 to 90 seconds. |
| Re-evaluation cadence | eBay’s official seller-standards evaluation runs monthly on the 20th. This card refreshes continuously. The card’s number leads the official evaluation by up to 30 days. |
| Reason | Direction | Why |
|---|---|---|
| Window length | Either | This card displays both 30D and 12M rolling rates. Seller Hub typically shows the 12M figure on the seller-level page; the rates can disagree by 0.05 to 0.20 points if recent performance has shifted. The 12M number is what eBay evaluates against; the 30D is a leading indicator. |
| De-duplication | Same on both | Both views de-duplicate transactions with multiple defect categories; counts of “defect events” can mislead. Always work with unique-transaction defect rate. |
| Refresh latency | Ours within 90s | A defect resolution that closes a case favourably can be reflected on Seller Hub within seconds; our card refreshes within 90 seconds. |
| Card | Expected relationship | What causes legitimate divergence |
|---|---|---|
amazon.account_health | Marketplace peer. Amazon’s Order Defect Rate (ODR) is structurally similar but capped at 1% (vs eBay’s 0.5% for TRS). | Different cap, different defect categories. eBay counts “Cases without seller resolution”; Amazon counts negative feedback, A-to-z claims, and chargebacks. Used for peer-benchmarking only. |
shopify.refund_rate | Loose comparison. Shopify’s refund rate is a count of refunded orders, distinct from a defect rate (a defect can occur without a refund, and a refund can occur without a defect). | Independent metrics with different meanings; treat as orthogonal, not reconcilable. |
Known limitations / merchant FAQs
What counts as a defect? Five categories: (1) cases closed by eBay without seller resolution, (2) seller-cancellation due to out-of-stock or “can’t ship”, (3) late shipment with no tracking uploaded on time, (4) INR cases auto-decided against the seller, (5) “Item Not as Described” returns the seller refused or escalated. eBay groups several of these under “service metrics”; the Sell Analytics API exposes them individually so the card can show per-category contributions. Why is the 0.5% cap so harsh? TRS is eBay’s premium-quality signal to buyers. A seller pushing 1% defects is making 1 in 100 buyers unhappy enough to escalate, that’s 10x the typical Amazon threshold for similar reasons. eBay’s threshold is intentionally tight to keep the TRS pool genuinely high-quality. How long does it take to lose / regain TRS based on defect rate? Lose: one monthly evaluation (the 20th of each month) where the rolling 12-month rate is above 0.5% triggers the demotion. Regain: the rolling 12-month rate must drop back below 0.5% by a subsequent evaluation. Because defects persist in the 12-month window, regaining usually takes 6 to 12 months unless the seller had a single bad month that drops out of the window. What’s the relationship between defect rate and the Health Score? Defect rate is one of four inputs to the Health Score, weighted at 25%. A defect-rate jump of 0.3 points (e.g. 0.4% to 0.7%) drops the Health Score by ~7 points all else equal. Does Best Offer affect defect rate? Indirectly. Best-Offer-resolved orders involve more buyer messaging before the sale, which builds expectation alignment and tends to reduce the per-order defect risk. Sellers who run aggressive auto-accept rules (no human dialogue) lose this protective effect, their Best-Offer orders defect at the same rate as cold BIN orders. My multi-site defect rate, where does the variation come from? ebay.co.uk typically runs higher defect rates than ebay.com because UK consumer-rights law (Consumer Rights Act 2015) gives buyers stronger refund leverage and a 14-day cooling-off period; UK buyers escalate faster. ebay.de runs the lowest because German buyers tend to message before disputing. Look at per-site rates and tune practices accordingly. Promoted Listings: do they affect defect rate? Indirectly via volume. Promoted-driven orders defect at similar per-order rates to organic, but if you scale Promoted on defect-prone clearance / refurbished SKUs, the absolute defect count rises. Action: monitor defect rate when scaling Promoted, and pause Promoted on any SKU whose defect rate is above the brand average. Why doesn’t defect rate match Shopify or Amazon? Different mechanisms. Shopify has no defect-rate equivalent; refunds and disputes go through the payment processor. Amazon’s Order Defect Rate (ODR) has a 1% cap and counts negative feedback, A-to-z claims, and chargebacks. The categories don’t overlap cleanly with eBay’s five defect types. Treat as orthogonal metrics, not reconcilable. Today’s defect rate jumped 0.2 points but I haven’t shipped anything new, why? Two usual causes: (1) eBay closed a case-without-seller-resolution that had been pending for days, that defect just got recorded retroactively against an order from weeks ago; (2) a buyer left negative feedback that eBay categorised as a defect. The 30D rate updates retroactively on the day the defect is recorded, not the day the order was placed. Watch Cases without Resolution for the full forensic detail. Action playbook when this card alerts (>0.5%):
- Open Cases without Resolution. Identify any active cases approaching auto-resolution; refund proactively to close favourably.
- Open Open INR Cases. Respond to every case under the 5-day threshold.
- Audit out-of-stock cancellation patterns from the last 30 days, recurring cancellation = inventory-sync issue, fix the warehouse-to-listing flow.
- Pull the per-SKU defect concentration via Vortex Mind. 80% of defects usually come from 20% of SKUs; pause those listings or fix descriptions / photos.
- If the 12-month rolling rate is also above 0.5%, prepare for the next monthly evaluation, communicate to ops the cliff-effect on payouts and FVF.