THE METHOD

Watch. Classify. Re-baseline.Tell first. Close.

The event arrives raw — a discharge scan at a hub, a blanked departure, a hold code. What must leave is a customer who already knows the new plan. Five moves close that gap, with your approval on everything outbound.

01 · WATCH

Every box on the sweep. Continuously.

The manual version of tracking is a morning ritual: a coordinator opens each carrier portal in turn, re-keys milestones into an Excel sheet, and hopes nothing moved since it saved. The desk replaces the ritual with a watch — polling carrier sources across ocean, air and express continuously, resolving each shipment the way freight is actually numbered (booking number, then BL, then container), and holding every milestone current from empty gate-out to empty return. Nothing depends on someone remembering to look.

ocean / air / express · continuous · portals + spreadsheets retired

02 · CLASSIFY

Anomalies become named exceptions.

A raw event feed is homework, not an answer. The desk reads the pattern and names the exception — rollover, blank sailing, customs hold, transshipment delay — with its reasoning shown, not just its verdict. Rollovers are the proof case: carriers rarely announce them, so the desk infers them — your box discharged at the transshipment hub but never loaded onto the booked vessel. That mismatch is a rollover whether anyone said so or not, and it’s caught when it happens, not when your customer’s trucker turns up against the old ETA.

ROLL · BLANK · HOLD · T/S · discharge/load mismatch → ROLL

03 · RE-BASELINE

The dead ETA is replaced everywhere.

An exception makes the old ETA fiction — and fiction is what customers book truckers against. The desk recomputes the arrival from what actually happened: the vessel the box actually loaded on, the new rotation, the exam clock still running. Then it identifies everyone whose plans hung on the old date — the consignee’s delivery slot, the notify party, the downstream leg — so one new baseline replaces the dead one across the whole file, and no two people are working from different truths.

new ETA computed · affected parties identified

04 · TELL FIRST

The update lands before the question.

Every affected customer gets a drafted update in their language, on their channel — email or WhatsApp — saying what happened, why, and what the new plan is. Cadence is configured per customer: some want every move, some a weekly digest. And it’s never a firehose — milestones that matter become updates (departed, arrived, out for delivery, exception, POD); internal noise doesn’t. Each draft waits for your approval until you dial that update type to auto. The desk earns autonomy the way a new hire does.

their language · their channel · your approval until auto

05 · CLOSE

The paper exists the day it’s earned.

Delivered isn’t done. Many customers’ payment clocks start only when the signed POD reaches their accounts team — and that paper has historically surfaced days to weeks after the freight did. The desk collects the proof of delivery as the delivery completes and attaches it to the shipment, so the document your invoice may be gated on exists on day one, not whenever it resurfaces. The file closes with its whole story attached: every event, every exception, every message sent.

POD attached at delivery · invoice-gating paper on day one

THE FIRST 14 DAYS

DAY 1–3

We scope the watch: your lanes, carriers and modes, plus the customer map — who gets updates, in which language, on which channel, at what cadence.

DAY 4–10

The desk shadows your live shipments — classifying exceptions and drafting updates your team grades against what actually happened. Nothing goes outbound.

DAY 11–14

Live under approval: real updates reach real customers after your sign-off — and you count the status queries that stopped arriving.

Put a lane on the watch