What Market Reliability measures, how on-time performance and ratings are defined, and how to read the numbers with confidence

Carrier selection is often made on gut feel, on what a carrier's sales team promises, or on spreadsheets that are months out of date. Carriers are not obliged to meet the transit times they advertise, and real schedule performance varies a lot from one service to another and from one month to the next. A carrier that holds its schedule on one trade lane can be the weakest option on another.

Market Reliability gives you the evidence: for any lane you choose, it shows how every carrier and every bookable service has actually performed against the schedules those carriers published, and Marina, the Portcast AI Market Reliability Agent, turns those numbers into a short written read with a recommendation.

<aside> 💡

Here are some questions Market Reliability helps you answer:

• Which carrier should I contract on Shanghai → Rotterdam, and on which of its services?
• Is the transit time a carrier advertises on this lane realistic, and how much buffer should I plan?
• Which service should I book for time-critical cargo on this lane?
• Are the carriers I already have contracts with better or worse than the rest of the market on my lanes?
• Is reliability on this lane improving or getting worse over the last months? </aside>

What does Market Reliability measure?

Market Reliability is lane based. You pick an origin port and a destination port, and the page analyses every carrier and every bookable service that connects them, scoring each one on how its completed sailings arrived against the schedule the carrier itself published.

The analysis is built from two things:

• Published schedules: the sailing schedules carriers publish for the lane, collected directly from the carriers and refreshed daily, so what you see matches what the carriers themselves announce.
• Confirmed arrivals: the actual arrival of each sailing at the destination port, confirmed from vessel positions and matched to the exact vessel and voyage the carrier scheduled.

Market Reliability measures the carriers' own schedule integrity, not the performance of your own shipments. For transit time analysis based on your tracked containers, see Using the Transit Times dashboard.

How is "on time" defined?

A sailing counts as on time when the vessel arrives at the destination port within two days either side of the arrival date the carrier was publishing around three weeks before departure, which is when most bookings are made.

That reference date is captured once and never revised. Carriers update their schedules continuously, and if reliability were measured against the latest ETA a carrier that keeps re-forecasting its arrival until it matches reality would look perfect. Freezing the reference at booking time means the score reflects what you could actually plan against when you booked.

Only sailings with a confirmed arrival are measured. Sailings that are still at sea or have not departed yet are counted as pending and do not affect the reliability figures until they complete.

Carriers and bookable services

The thing you book is not a carrier, it is a service: a fixed loop of vessels between a set of ports, sold under a service code. Market Reliability therefore scores the lane at two levels.

• Bookable services: each end-to-end option you can actually book on the lane, identified by its service code and the carrier selling it. One vessel is often sold by several carriers under different codes (vessel sharing agreements and alliances), so the same sailing can appear as several bookable rows with identical figures, one per carrier. Services that reach the destination via a transhipment hub are shown with the hub in the Via column, and multi-leg journeys carry the codes of each leg joined together (for example TPM+WC5).
• Carriers: the overall performance of each carrier on the lane, pooling all the measured sailings of all its services. This is the level to use for contract and tender discussions; the service level is the one to use when choosing what to book.

The metrics

Every row in the service and carrier tables carries the same handful of metrics: