Most sourcing advice is written for merchants and talks about cost per unit. This is written for the people who have to make the calendar work when the cost per unit changes in the middle of a season.
When tariffs move, the boardroom question is always the same: how fast can we move production somewhere else? The honest answer is slower than anyone wants, and for a reason most people guess wrong.
Realistic timelines run around ninety days for a replenishment programme on an existing style, and six to nine months for new or complex styles. That gap surprises people.
The limiting factor is almost never factory capacity. Capacity is usually available. What takes the time is fabric development, quality approval, and handing over tech packs to a factory that has never made your product. A new factory has to source or develop your fabric, hit your hand feel and colour standard, and interpret your specs correctly. None of that compresses just because a tariff changed.
The operational consequence: brands that pre-qualify two backup countries per major category, before they need them, move at ninety days. Brands that start looking when the tariff lands move at nine months and eat the difference.
A factory sitting inside a preference country does not mean your garment qualifies. This is where brands lose duty savings they thought they had banked.
Each agreement sets a different transformation test, and they are not interchangeable. USMCA requires yarn-forward, meaning the yarn itself has to originate in the region. The EU's Vietnam agreement requires fabric-forward. The UK's DCTS requires single-stage transformation plus direct shipment. A garment cut and sewn in a qualifying country can still fail if the fabric came from somewhere else, or if the documentation is incomplete.
This is not a paperwork inconvenience. Customs authorities can claw back the duty you did not pay, and penalties on under-collected duty have run as high as forty percent where documentation fails on audit. That lands months or years after the goods sold, against margin you already recognised.
The practical control is simple and rarely done: verify origin documentation before claiming a preferential rate, not after. If you cannot produce the yarn or fabric origin evidence on demand, you do not have a preference claim. You have an exposure.
Which of your styles sit close to the line on yarn-forward eligibility? Most brands have a handful that qualify only because of one fabric sourcing decision nobody has revisited in two years. Those are the ones that fail an audit.
The Uyghur Forced Labor Prevention Act creates a rebuttable presumption: goods with any Xinjiang link are assumed to involve forced labour and are barred unless you can prove otherwise. The burden sits with you, at the border, in real time.
Cotton is the exposure most apparel brands underestimate, because the risk is not at your factory. It is upstream, at the gin and the spinner, several tiers beyond the supplier you actually talk to. A Vietnamese factory can be entirely legitimate and still receive yarn from a source you cannot document.
Enforcement is not theoretical. US Customs detained roughly 1.4 billion dollars of apparel in 2025 under forced labour and origin enforcement. A detained container does not just cost duty. It misses the delivery window, which triggers retailer chargebacks, which is where an upstream sourcing problem turns into an operations problem on your desk.
You need documentation tracing cotton back through the supply chain, not a supplier attestation that they comply. The distinction matters at the border. Worth checking whether your garment labels and production records already double as traceability evidence, because in many cases they partly do and nobody has organised them for that purpose.
Any guide quoting current tariff rates is out of date within months. The last two years have made that obvious.
Rates on major sourcing countries have swung dramatically and then partly reversed. The de minimis exemption that let sub-800-dollar parcels enter duty free was suspended, which reshaped the economics of direct-to-consumer shipping into the US overnight. Measures introduced as temporary carry expiry dates that then become planning problems of their own.
The scale is not marginal. Lululemon disclosed a tariff bill in the hundreds of millions and rising year over year. For a mid-sized brand the proportional impact is the same even if the absolute number is smaller.
That last one is worth checking today. Many brands discover mid-season that their supplier agreement is silent on it, which means the conversation happens under pressure with goods already in production.
Most of the above fails operationally for the same reason: the information exists somewhere in the business but not in one place, and not in a form anyone can act on quickly.
Landed cost is the clearest example. If freight, duty and broker fees are not carried against the style, your margin reporting is fiction and every sourcing decision is made on incomplete numbers. Most systems can hold landed cost. Comparatively few brands have actually configured it, because it requires someone to own the data rather than the feature to exist.
Origin and traceability documentation has the same problem. It usually lives in email threads and supplier folders rather than attached to the style record, which means an audit becomes an archaeology exercise under time pressure.
And when a preference claim fails or a container is detained, the cost surfaces as a chargeback and a missed delivery window. Which is why a sourcing problem shows up first on a retailer scorecard, and gets diagnosed as a logistics problem.
We map where your origin evidence, landed cost and supplier terms actually sit, and what breaks when policy moves. If your setup is already sound, we will tell you that.
Book a call