At a glance
Absolute count of Royal Mail consignments that arrived after their service-code aim date in the trailing 7 days. The numerator behind the OTD percentage; the workload number for customer-service teams.
| What it counts | COUNT(shipments WHERE delivered_at > aim_delivery_date) over the rolling 7-day window. Each consignment scores once against its own service-code aim. |
| API endpoint | GET /shipping/v3/shipments (Royal Mail Shipping API v3). Reads serviceCode, aimDeliveryDate, actualDeliveryDate. |
| Service-tier scope | Tracked services only (Tracked 24, Tracked 48, Special Delivery, International Tracked). Untracked 1st / 2nd class are excluded because there is no delivered scan to compare against the aim. |
| Tracked vs untracked split | Untracked late deliveries are estimated separately on Failed Deliveries using customer-service ticket signals; the count on this card is tracked-only. |
| Return-leg inclusion | Outbound only. Tracked Returns and Returns Portal consignments are filtered out via direction = inbound. |
| Geographic scope | UK domestic AND International Tracked & Signed. Highlands & Islands, BFPO, and Channel Islands lanes are included; international untracked is excluded. |
| In-flight handling | Consignments currently in transit (actualDeliveryDate IS NULL and aimDeliveryDate >= now) are excluded. Consignments past their aim and still in transit (aimDeliveryDate < now, no delivered scan) are counted as late. |
| Strikes / industrial action | Counted as late. Use the operations playbook to suppress alerts during known stoppage dates rather than altering the count. |
| Time window | 7D (rolling 7 days). Daily readings are noisy below 200 consignments per day. |
| Alert trigger | >5% of total (the late count exceeds 5 percent of total tracked consignments in the period). For a 4,000-parcel-week merchant this is 200+ late deliveries per week. |
| Roles | owner, operations |
Calculation
Calculated automatically from your Royal Mail data. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.Worked example
Same UK DTC homewares brand from OTD Rate, Manchester-based, 60 percent Royal Mail Tracked 48 / 40 percent Evri ParcelShop. Reading taken at 09:00 BST on 12 Mar 26 for the trailing 7 days (06 Mar 26 to 12 Mar 26). Royal Mail leg only:| Service code | Consignments (7D) | Late (delivered after aim) | Late % |
|---|---|---|---|
| Tracked 48 (TPN48) | 2,310 | 89 | 3.9% |
| Tracked 24 (TPN24) | 380 | 23 | 6.1% |
| Special Delivery 1pm (SD1) | 65 | 1 | 1.5% |
| All RM tracked (this card) | 2,755 | 113 | 4.1% |
>5% of total is not tripped at the aggregate (4.1 percent) but Tracked 24 alone is at 6.1 percent. Five things to notice:
- The percentage felt fine on the OTD card; the count is the workload. 113 late deliveries equals roughly 113 customer-service tickets currently in flight or about to land. At an estimated 8 minutes per WISMO ticket, that is 15 staff-hours of work generated this week by carrier delays alone. The CS team needs to know.
- Tracked 24 is the actionable signal. It is breaching its own contract floor; raise a service-credit ticket with the Royal Mail account team using the Click & Drop performance report as evidence. Tracked 48 within tolerance is normal; Tracked 24 above 5 percent is not.
- Compare to the same week in 2022. The CWU industrial action of 26 to 31 August 2022 caused this same brand’s weekly late count to spike from a normal 80 to 130 range to over 600 late shipments in one week. The “rate suddenly degraded” debug case is almost always either industrial action or a regional sortation centre incident. Check the Royal Mail Service Update page first, the strike calendar second.
- The 113 count under-states the customer-felt impact. Untracked 1st class is invisible here, but a typical merchant has another 30 to 50 percent of volume on untracked, of which 5 to 10 percent is late by perception. Real customer-impact is roughly 1.5x to 2x the tracked late count. Use this number for ops planning, not customer-experience modelling.
- The Special Delivery 1 late is recoverable. Royal Mail SLA-credits Special Delivery automatically if it misses its 1pm guarantee, the merchant should reclaim £8.10 to £10.00 per missed parcel. For 1 parcel that is rounding; for a brand sending 200 SDs a week with 1 percent miss rate, it is £1,500 / year of leakage.
Sibling cards merchants should reference together
Late shipment count is the workload metric. Pair it with these to plan response:| Card | Why pair it with Late Shipments | What the combination tells you |
|---|---|---|
| On-Time Delivery Rate | The percentage view of the same metric. | The percentage tells you how the network is performing; the count tells you how much work the CS team has tomorrow. Track both. |
| Exception Rate | What fraction of consignments threw a tracking exception. | A rising exception rate predicts a rising late count at 24 to 48 hours. |
| Open Claims | How many late deliveries escalated into claims. | Typical conversion is 10 to 25 percent: 100 late shipments produces 15 to 25 P58 claims. |
| Tracked24 Service Day Promise | Premium-tier-only late count is service-credit recoverable. | Late Tracked 24 deliveries should be reclaimed from Royal Mail; budget for ~80 percent recovery rate if the claim is filed within 28 days. |
| Failed Deliveries | Untracked failure proxy. | Adds back the invisible-on-this-card untracked tail, gives a closer “true” late figure. |
Cross-connector: shopify.refund_rate | Downstream impact. Late deliveries drive refund tickets at 7 to 14 day lag. | A weekly late-count spike of 100+ predicts a 0.5 to 1 percentage-point refund-rate rise the following fortnight. |
Cross-connector: hermes_evri.her_late_shipments_count | Adjacent carrier when the merchant runs both. | Compare ratios, not raw counts, the carriers handle different volumes. A spike on one carrier and not the other points to carrier-specific cause. |
Cross-connector: shipbob.sb_otd_rate | When ShipBob is the warehouse and Royal Mail is last-mile. | If ShipBob ships on time but Royal Mail late count spikes, the issue is at the carrier handover, not the warehouse. |
Reconciling against the vendor’s own dashboard
Where to look in Royal Mail’s own portal: Royal Mail Click & Drop → Reports → Delivery Performance → Late Deliveries (filter), or for Business Account holders Royal Mail Business Account → Performance → Service Performance Report → Missed-aim consignment list. The Business Account view is the official record used for service-credit claims under the Royal Mail compensation scheme. The closest like-for-like view is All Tracked Services, Last 7 Days, Outbound Only, Status: Late. Why our number may legitimately differ from Royal Mail’s report:| Reason | Direction | Why |
|---|---|---|
| Timezone (BST vs UTC) | Boundary days off | Royal Mail’s portal uses UK local (BST in summer, GMT in winter). The card stores in UTC. For a 7-day window, boundary-day shifts cause 5 to 30 consignments to land in different days; the weekly total typically agrees within 1 percent. |
| Tracking-event ingestion lag | Ours can lag | Royal Mail’s tracking-events feed pushes scans in 30-minute to 6-hour batches. A consignment delivered 4 hours ago may not show as delivered in our index yet; conservative classification means we err on the side of “still in transit” rather than late. |
| Service-tier reclassification mid-period | Either | Mid-period upgrades (Tracked 48 to Tracked 24) split the population in the portal; the card uses label-time service code. |
| Peak-period sampling | Either | December peak: Royal Mail’s reporting cadence may show partial data for the most recent 5 to 7 days. The card always shows confirmed deliveries plus past-aim in-transit. |
Cross-connector reconciliation against shipbob.sb_otd_rate | Different population | When ShipBob is warehouse and Royal Mail is last-mile, ShipBob’s late count is warehouse-floor + carrier; this card is carrier-only. A late ShipBob shipment is not necessarily a late Royal Mail consignment. |
| Card | Expected relationship | Causes of legitimate divergence |
|---|---|---|
shopify.unfulfilled_orders | Upstream input. | Webhook delivery failures, manual fulfilment delays, B2B / pre-order. |
shipbob.sb_otd_rate | Adjacent 3PL when ShipBob handles warehouse. | Different scoring (warehouse + carrier vs carrier-only). |
Known limitations / merchant FAQs
Royal Mail vs Evri, why does my Evri late count look smaller even though OTD is similar? Carrier mix. If you send 60 percent of volume via Royal Mail and 40 percent via Evri, the Royal Mail count will be 1.5x the Evri count even at the same OTD percentage. Compare the late-rate ratio (count divided by total per carrier), not the raw count. Read this card alongsidehermes_evri.her_late_shipments_count and apply the per-carrier divisor.
What do I do during a Royal Mail strike?
The CWU industrial action of Aug 2022 to Apr 2023 caused the late count to spike 4x to 8x for many merchants on stoppage weeks. Three actions: (1) suppress alerts on this card for the announced stoppage dates so the CS team is not paged for a known event, (2) shift sensitive volume to Evri or DPD for the affected days, (3) re-baseline thresholds temporarily and reset when service resumes. Do not permanently raise the threshold; the carrier should be held to its baseline once industrial action ends.
Tracked vs untracked, why is untracked invisible here?
Untracked Royal Mail (1st class, 2nd class) does not generate a delivered scan. The carrier records the consignment as “deemed delivered” after a notional aim window (3 working days for 1st class, 5 for 2nd) but never confirms. Late untracked is detectable only via customer-service ticket signals or absence-of-delivery surveys; it is on Failed Deliveries using a separate methodology. Treat the two views as complementary.
My merchant has Tracked 48 default but pushes Tracked 24 at checkout for £2 extra. How should I read the count?
Per service code, not aggregate. Tracked 24 has a tighter aim and a separately negotiated SLA contract. A late Tracked 24 is more expensive to your customer relationship and more recoverable from Royal Mail. Use Tracked24 Service Day Promise for the premium leg in isolation, this card for the combined workload.
Why does the count bounce around day-by-day even when OTD is steady?
Volume seasonality. Tuesday and Wednesday have 30 to 40 percent higher despatch volume than Monday or Friday for typical UK DTC. A steady 96 percent OTD on Tuesday produces 1.4x the late count of the same OTD on Friday. Use the 7-day rolling view, not daily.
What is the playbook when the count crosses 5 percent of total?
In order: (1) check Royal Mail service updates for known network incidents or industrial action, (2) check Exception Rate for upstream label-data issues (address-invalid, missing-postcode), (3) check OTD by Route for regional concentration, (4) raise a ticket with the Royal Mail account team if (1) to (3) do not explain it, attach the Click & Drop performance report.
How do I plan for Christmas peak?
Expect this count to roughly double from 1 December to 23 December. Last-recommended-posting dates are published mid-October each year by Royal Mail; treat them as a hard cliff in checkout copy. Three actions: (1) tighten the checkout deadline messaging, (2) shift budget volume to Evri / DPD for non-time-sensitive items, (3) hold Special Delivery 1pm capacity for premium gift items by negotiating with the account team in October.
Why does Shopify show “shipped” but Royal Mail tracking shows nothing for the parcel I am about to count as late?
Two-stage handover. Shopify marks shipped on label-print (fulfilment.created); Royal Mail’s first scan is at collection or Post Office handover. Typical lag is 2 to 24 hours. If 48 hours pass without a first scan, the parcel was probably never collected, check Collection vs Post-Office Handover Failure Rate. The card classifies a never-scanned parcel as in-transit until past its aim, then late, even if it was never actually handed over to the carrier.