At 8:15 on a Thursday morning a trade compliance manager in Toronto opens a forwarded email with a Federal Register link in it. Four days from now, a category of goods her company ships into the United States every week stops being importable at all. Not dutiable at 50 percent. Not quota-limited. Banned. Her first question is not strategic, it is arithmetic: how many of our shipments does this touch, where are they right now, and which ones can still be entered before the cutoff. Nobody in the building can answer her before lunch.
Why this clock works differently
On September 14 the White House published proclamations under Section 338 of the Tariff Act of 1930 excluding certain Canadian products from importation into the United States, effective for goods imported on or after 12:01 a.m. eastern time on September 29. A separate proclamation the same day changed which products sit inside the existing duty scope, adding some and dropping others with effect from September 15. Going the other direction, Canada's counter-tariffs of up to 50 percent on roughly 700 categories of US goods took effect on September 8.
Three dates inside three weeks, each attached to a different product list. Now look at what the effective date actually attaches to. It is not the day you booked the container, and it is not the day the vessel sailed. It is the moment the goods are imported. A shipment that left Ontario two weeks ago and a shipment that leaves tomorrow can land on opposite sides of the same rule, depending on when the entry gets filed and released. That is an operational variable sitting inside your own systems, and it is the one you can still influence this week.
The question your systems were not built to answer
The exposure question sounds simple and cuts across four systems that have never been introduced to each other.
The ERP holds the purchase order line, and with it the commodity, the value and, if you are lucky, a classification. The TMS or the forwarder portal holds the container or the air waybill, the carrier and an ETA. The WMS holds what already landed and is sitting in a distribution center. The broker's platform holds entry status, which is the field the rule actually turns on. Four systems, four different primary keys, no join.
So the morning goes the way it always goes. Someone exports three spreadsheets. Someone else matches container numbers against purchase order lines by hand. By the time the list exists it is a day old, half the ETAs have moved, and nobody is confident it is complete.
This is the join a supply chain command center exists to make. When the order line, the shipment, the customs milestone and the inventory record hang off one normalized object, "every unit of these HS codes, Canadian origin, not yet entered, arriving before Monday" is a filter rather than a project. Orkestra sits above the ERP, TMS and WMS instead of replacing any of them, so the commodity data from procurement and the position data from transport are already on the same record when a rule lands with a date on it.
Scope from the entry, not from the shipment
Work the list in this order.
Start with codes, not descriptions. Pull the rule's product list and resolve it to HS classifications at eight or ten digits. Shipment-level descriptions will not do this job. "Auto parts" and "foodstuffs" are what shows up on a bill of lading, and neither tells you whether a line is in scope.
Match at the order line, not the shipment. A single container routinely holds items on both sides of a rule. Scoping at container level either overstates your exposure and triggers a panic, or understates it and misses the lines that matter.
Filter on origin, not on where it shipped from. Country of origin and country of loading are different fields and diverge constantly. Goods transhipped through Canada may not be Canadian origin at all. Goods manufactured in Canada and already sitting in a Buffalo warehouse still are.
Attach an entry state to every unit. Not yet arrived, arrived but not entered, entered and released, or already in stock. That state, not the geography, is what decides which pile a shipment belongs in. If your arrival milestone and your release milestone are the same field, you cannot tell those states apart, and the distinction is the whole exercise.
Sort the result into three piles
Can still make the cutoff. These need broker lead time more than they need transit speed. Get documents over early, file ahead of arrival where the process allows it, and put a named person on each one. A day of paperwork lead time is worth more this week than a day of ocean transit.
Will miss it. Decide now whether to divert, hold in bond, re-consign to another market, or sell where the goods already are. Every one of those options gets cheaper the earlier it is taken and most of them stop being available once the box is at the border.
Already landed and cleared. This pile is safe from the rule and it is also, as of the cutoff, the only inventory you have for those SKUs. Tell demand planning and customer service today rather than when the first order fails to allocate.
Most teams work the first pile hard and forget the third entirely. The customer-facing damage from a trade rule almost never arrives on the day the rule takes effect. It arrives six weeks later as a stockout nobody connected back to it.
The five fields that decide how fast you can answer
Origin at line level. HS classification carried on the order rather than guessed at the border. Entry state. Current position with an ETA you actually trust. And an owner, so each line has someone accountable for the next action.
If your classifications live only in the broker's system, you cannot scope your own exposure without asking them, and this week they are answering the same question for every other client on their book. Pulling classification and origin upstream into the order record is the highest-return data fix available for tariff volatility, and it is the same fix that shortens routine customs clearance in the months when nothing is on fire. Getting to one joined record is its own piece of work: building a supply chain data model for end-to-end visibility covers what that takes.
The payoff shows up well beyond trade policy. DBW Advanced Fiber moved off spreadsheets to real-time EU to North America visibility on Orkestra and cut supply chain costs by 18 percent. Answering a tariff question in an afternoon is a side effect of that same joined record, not a separate capability you buy.
Working the list once you have it
A scoped list is not a plan. Rank it by what breaks downstream rather than by unit value, because the expensive shipment and the one that stops a line are rarely the same shipment. The ranking logic is the same one you should already be using when everything is late at once.
Then make the follow-up survive the week. Every line needs an owner, a deadline and an escalation path, because a list assembled on Thursday and left in a spreadsheet is worthless by Tuesday. Running that as structured workflow with named ownership, rather than as a thread of replies, is what turns a scoping exercise into shipments that actually moved.
Before the next one
September 29 will pass. The pattern will not. Rules with short fuses and specific product lists are now a standing feature of operating across borders, whether the structural answer you are pursuing is regional sourcing or simply a deeper bench of alternatives.
So turn this week's query into a saved view. If you can ask your own systems, on any Tuesday, "what do we have in flight of this commodity, from this origin, not yet entered," then the next proclamation costs you an afternoon instead of a quarter. That standing answer is what Orkestra is for, and it is worth having in place before you need it. Book a demo and bring the question you could not answer this morning.
