BENCHMARKS
The schedule is only two-thirds true.
The reference numbers behind proactive tracking — 2026 schedule reliability month by month, blank-sailing ranges, what the status-question burden costs in the only industries that measured it, and rollover history with its vintage clearly labelled. Sources named on every table, flags applied in-line, and an honest section on the benchmark nobody has published.
SOURCES: SEA-INTELLIGENCE · DREWRY · RETAIL CX BENCHMARKS · UPDATED JULY 2026
Schedule reliability, month by month
Sea-Intelligence publishes the industry’s headline on-time metric monthly. The 2026 readings so far say the same thing five different ways: roughly one vessel in three arrives late — which is why an ETA is an opening position, not a promise.
| Month (2026) | Schedule reliability | Late arrivals (derived) |
|---|---|---|
| Jan | 62.4% | 37.6% of arrivals missed schedule |
| Feb | 59.0% | 41.0% — the weakest month so far |
| Mar | 62.2% | 37.8% of arrivals missed schedule |
| Apr | 62.4% | 37.6% of arrivals missed schedule |
| May | 64.7% | 35.3% — the strongest month so far |
SOURCE: SEA-INTELLIGENCE GLOBAL LINER PERFORMANCE (PRESS ROOM) · MAY FIGURE VIA CONTAINER NEWS — PAYWALLED PRIMARY
When vessels do run late, the average delay across these months sits around 5.2–5.5 days. One flag before you republish: the per-month average-delay figures sit behind Sea-Intelligence’s paywalled primary — verify them against the report itself before citing month-specific delay numbers anywhere that matters.
Blank sailings
A blank sailing is a scheduled departure the carrier cancelled — and for anything booked on it, an ETA that needs re-baselining and a customer who needs telling. Drewry tracks cancellations weekly.
| Figure | What it measures | Source |
|---|---|---|
| 3–18% | of scheduled departures blanked across 2026 windows — the range moves week to week as carriers manage capacity | Drewry Cancelled Sailings Tracker (weekly) |
| 46–63% | the Transpacific's share of all cancelled sailings in tracked 2026 windows — the lane that absorbs the most blanks | Drewry Cancelled Sailings Tracker (weekly) |
SOURCE: DREWRY CANCELLED SAILINGS TRACKER · RANGES REFLECT 2026 WEEKLY WINDOWS
The operational point: the cancellation notice, when it arrives, goes to the booking party — your customer finds out from you, or finds out late.
The service burden
Freight has not benchmarked its own status-question load, so the citable numbers come from adjacent territory. Read the source column before borrowing any of these: the retail rows are context, not freight data, and the check-call rows carry a provenance flag.
| Figure | What it measures | Source & flag |
|---|---|---|
| 20–40% | of e-commerce support tickets are WISMO — “where is my order?” | Retail CX benchmarks — retail context, not freight-specific |
| ~$5 | typical cost per support contact in the same retail benchmarks | Retail CX benchmarks — same context label applies |
| 40% | of an operator's time reportedly spent on check calls | Vendor-published (Debales), attributing McKinsey 2024 — primary not located; directional only |
| 60–80 | check calls per coordinator per day, per the same claim | Same vendor-attributed figure — verify before citing |
| 46% | of shippers say they have left a forwarder for one with better technology | Logixboard / FreightWaves survey — vendor-commissioned; the strongest switching stat available |
SOURCES NAMED PER ROW · RETAIL CONTEXT AND VENDOR-ATTRIBUTED FIGURES FLAGGED IN-LINE
The pattern behind the flags is consistent, though: the status question is a measurable cost center everywhere anyone has measured it — and the freight version arrives on WhatsApp, marked urgent, forty times a day. The vocabulary for all of it lives in the glossary.
Rollover, historically
The rollover rates that circulate online are mostly crisis-era. Use them for what they prove — the mechanism, not the current rate.
| Figure | What it measures | Source |
|---|---|---|
| 28.5–39% | rollover rates recorded at major transshipment hubs across 2020–21 — pandemic-crunch peaks, not a steady state | Ocean Insights (2020–21) — flagged: crisis-era vintage |
SOURCE: OCEAN INSIGHTS · 2020–21 FIGURES FLAGGED AS CRISIS-ERA VINTAGE
The durable lesson survives the vintage: rollovers are discovered, not announced. Whatever the rate is this quarter, no carrier message tells you your box stayed on the quay — detection is on whoever is watching the discharge/load pairs.
The number that doesn’t exist
No public benchmark exists for the figure this category actually turns on: hours per week a forwarder’s customer-service team spends answering status queries. We searched — the WISMO numbers are retail, the check-call numbers are vendor-attributed, and nobody has measured the forwarder’s desk. Our survey across live forwarder operations will publish it: hours per week on “where’s my container?”, by team size and mode. Until then, we won’t pretend a number exists.
Turn the benchmarks into your number.
The calculator applies these rates to your own volumes — how many late arrivals, how many blanked sailings, how many hours of status queries the watch takes off your desk.
Open the calculator