At a glance
The early-warning alert that catches a slow, dangerous drift in account-health metrics before the headline status flips. Where Account Health Status tells you the account has already moved to At Risk, this card fires when Order Defect Rate or Late Shipment Rate is creeping upward week on week, the gradual slide that, left unchecked, becomes a suspension. It is the difference between fixing a trend and firefighting a crisis.
| What it counts | A week-on-week change detector on the key account-health metrics, chiefly Order Defect Rate (ODR) and Late Shipment Rate (LSR). It surfaces metrics that are rising, even when each is still inside Amazon’s pass threshold. |
| Why a separate card | The status gauge is binary-ish: Healthy until a threshold breaks. This card watches the slope, so you act while the number is still green but heading the wrong way. |
| Fulfilment scope | Account-wide, but the creep is almost always FBM-driven, because FBM is where the seller controls shipping and service. FBA’s shipping metrics are largely Amazon’s responsibility. |
| Typical drivers | A courier slipping on delivery times, a warehouse falling behind on handling time, a defective batch generating returns and claims, or a seasonal volume spike outrunning the fulfilment team. |
| Relationship to the status | This is the leading indicator; the status gauge is the lagging result. Catching the creep here means the status never has to flip. |
| Refunds / claims | A rising trickle of refunds, negative feedback, or A-to-z claims often shows up here first, as ODR ticks up a fraction of a point each week. |
| Time window | 7D (seven-day rolling, compared week on week to detect the slope) |
| Alert trigger | ODR or LSR up >0.2pp WoW. A rise of more than 0.2 percentage points week on week in either metric flips the card and notifies owner and operations. |
| Roles | owner, operations |
Calculation
Calculated automatically from your Amazon Seller Central 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 kitchenware seller, mostly FBM on amazon.co.uk. Three consecutive weekly snapshots leading to 21 Mar 26.| Week ending | Order Defect Rate | Late Shipment Rate | Status gauge |
|---|---|---|---|
| 07 Mar 26 | 0.45% | 2.1% | Healthy |
| 14 Mar 26 | 0.61% | 2.9% | Healthy |
| 21 Mar 26 | 0.84% | 3.8% | Healthy (just) |
- The status gauge is still green, but the trend is alarming. ODR is under the 1% target and LSR is under the 4% target, so Account Health Status shows Healthy. Yet both metrics nearly doubled in two weeks. The creep alert exists to flag exactly this, momentum the binary status cannot show.
- At this slope, the account breaches within a week or two. LSR went from 2.1% to 3.8% in a fortnight; one more bad week pushes it past 4% and the status flips to At Risk. Acting now is far cheaper than recovering a suspended account.
- Both metrics moving together points at one root cause. A simultaneous rise in defects and late shipments usually means a fulfilment bottleneck, a courier problem or a warehouse falling behind, rather than two unrelated issues. That makes diagnosis faster.
- This is the card that buys you time. Catching the slope here means a quiet operational fix (tighten handling time, switch courier, pause the worst SKUs) instead of a Plan of Action and an appeal later. The whole value is in the lead time.
Sibling cards merchants should reference together
Creep is a leading indicator. Pair it with the metrics it is watching and the status it is protecting:| Card | Why pair it with Account Health Creep |
|---|---|
| Account Health Status | The lagging result this card protects. The creep alert is designed to fire before the status gauge flips amber. |
| Order Defect Rate | One of the two metrics whose slope this card tracks. Check it to see the absolute level behind the trend. |
| Late Shipment Rate | The other tracked metric, and usually the faster-moving one. A climbing LSR is the most common creep driver. |
| A-to-z Guarantee Claims (open) | Granted claims feed ODR. A rising open-claims count is often the source of an ODR creep. |
| Pre-Fulfilment Cancel Rate | Cancellations after payment are a defect that can push ODR up gradually. Watch it alongside the creep. |
| Negative Feedback (30d) | Rising negative feedback often precedes an ODR creep by a week or two. A useful upstream signal. |
Reconciling against Amazon Seller Central
Where to look in Seller Central:Account Health → the Customer Service Performance and Shipping Performance panels. Amazon shows the current ODR and LSR values there, but it does not present an explicit week-on-week slope. To reconcile the creep, note the ODR / LSR values across consecutive visits, the deltas are what this card computes for you.Because Amazon shows the level and not the trend, the creep card is doing arithmetic Amazon does not display. Reconcile by confirming the underlying ODR and LSR values match; the week-on-week change is derived from those. Timing and reporting-lag table:
| Topic | Detail |
|---|---|
| Timezone | Amazon’s metric windows use the marketplace’s local timezone. Vortex IQ aligns the weekly comparison to consistent boundaries; a metric near a window edge can read slightly differently. |
| Rolling windows | ODR and LSR are themselves rolling metrics. A week-on-week comparison of rolling values smooths single-day noise, which is intentional, the card is looking for sustained drift, not one bad day. |
| Refresh cadence | The underlying metrics update on Amazon’s rolling schedule; the creep is recomputed each cycle so the slope stays current. |
| FBA vs FBM | FBM shipping metrics move fastest and drive most creep alerts. FBA delivery performance is largely Amazon’s responsibility and rarely the source of a creep. |
| Reason | Direction | Why |
|---|---|---|
| Derived vs displayed | Trend not shown on Amazon | Amazon shows the ODR / LSR level, not the week-on-week slope. The creep figure is computed by Vortex IQ from those levels, so there is no single Amazon number to match, reconcile the inputs instead. |
| Window alignment | Small deltas near boundaries | The exact week boundary affects the computed change. A metric that moved on the boundary day can shift the reported slope slightly. |
| Rolling smoothing | Either direction | Comparing rolling metrics dampens spikes. A sharp one-day event may show as a modest creep rather than a jump. |
| Card | Expected relationship | What causes legitimate divergence |
|---|---|---|
ebay.seller-standards-drop-alert | Marketplace peer. eBay’s standards drop alert is the equivalent early-warning on the eBay side, though eBay evaluates on a monthly cadence rather than continuously. | Independent systems and cadences. A creep on Amazon does not imply movement on eBay unless a shared fulfilment problem is behind both. |
shopify.return-rate | Shared-cause signal. A defective batch or courier failure can drive a creep on Amazon and a return-rate rise on Shopify at the same time. | Separate populations; correlated only when the root cause is shared fulfilment or product quality. |