Last updated: 28 July 2026. Trade measures and country-level duty rates are moving quickly at the moment; nothing here is legal, customs or tax advice, and no rate should be modelled from this page rather than from your customs broker.
China+1 is a genuine answer for some categories and an expensive rearrangement for others. Which one it is for you turns on where your inputs actually originate, what it costs to duplicate your tooling, and whether the alternative country makes what you buy to the standard you need. Below is how we work that out — including the cases where the honest recommendation costs us the work.
The question most people skip
Moving assembly doesn't move the supply chain.
The common outcome is production relocated to a new country while the components still ship from China. That adds a border, a lead time and a coordination burden without removing the exposure — and in some cases it makes tariff classification harder rather than easier, because substantial transformation has to be demonstrable, not just asserted.
It is also the part that is easiest to get wrong quietly. A supplier in a new country can be entirely honest and still be buying sub-assemblies, motors, fasteners, textiles or electronics from the same Chinese industrial cluster you were trying to step away from. Nobody is hiding it. It simply never comes up, because you asked where the factory is rather than where the parts are.
Before any relocation decision: trace where your inputs actually originate, not where your assembler is. If you cannot name the origin of your top components by value, you are not yet in a position to price the move — and any saving in the business case is provisional.
A note on rates
We're not going to publish a country comparison table.
You will find plenty of pages ranking alternative sourcing countries by duty rate and wage level. Most of them were true on the day they were written. Trade policy has been unusually unstable: the Section 122 surcharge lapsed on 24 July 2026 at the end of its statutory limit, tariffs imposed under IEEPA were held unlawful in Learning Resources v. United States on 20 February 2026, and Section 301 duties continue under a separate authority with no expiry attached. Country-level headline numbers move underneath a relocation plan that takes a year to execute.
So we keep this structural. Evaluate the things that change slowly — whether the industry exists, whether the component ecosystem is local, what tooling costs to duplicate, whether the supplier can hold your tolerances — and treat the duty differential as a variable you re-check with your broker rather than a foundation you build on. If you want the current arithmetic on your own lines, that is what our landed-cost re-model is for.
When it's right
Three conditions that usually have to hold together.
- The product suits it. Labour-intensive, low-tooling, low-precision products where a competent alternative industry already exists. If the making of your product is mostly hands and time rather than dies, jigs and process control, it travels.
- The exposure is product-specific. Categories facing product-specific trade measures rather than a general rate have a clearer case, because the thing you are moving away from is attached to your goods and not to the whole flow.
- Your volume matters to the new supplier. Volumes large enough to make you a priority customer somewhere new. A small order in an unfamiliar country buys you the same weak position you were trying to escape — that is the same failure described in single-supplier risk, relocated rather than solved.
When it isn't
Five reasons the answer is “not this category.”
None of these are arguments against diversification in general. They are the specific conditions under which a move consumes a year of management attention and returns a more fragile supply chain than the one you started with.
The cases where we say don't
High tooling costs you'd duplicate. Moulds, dies and fixtures do not follow you for free, and running two sets means paying for two sets of maintenance and two sets of first articles.
Deep component ecosystems that only exist in China. If your bill of materials draws on a cluster where a dozen sub-suppliers sit within an hour of each other, an assembler elsewhere is importing that cluster on your behalf.
Precision or process-dependent products. Where yield depends on accumulated process knowledge, a new plant starts at the bottom of the learning curve and your first year of output pays for it.
Volumes too small to matter to a new supplier. Minimums are higher, attention is lower, and you have no leverage when something goes wrong.
Timelines under a year. Qualification alone typically takes longer, and a rushed qualification is how a diversification project becomes a quality problem.
How to test it
Run a dual-source trial before you commit.
The decision does not have to be made on paper. A properly structured trial gives you a real answer for the cost of one production run, and it is the same sequence whether the second source is in another country or another Chinese province.
- Pick one SKU, not the range. Choose a product that is representative but not your highest-risk seller. You are testing a supplier and a country, not betting your best line on them.
- Write the spec as if the reader has never seen the product. Dimensions, tolerances, materials, finish, packaging, labelling, and the defects you will reject for. Every ambiguity your incumbent silently resolves from memory has to become text.
- Retain a golden sample from the incumbent. Sealed, dated, and used as the reference for everything that follows. Comparing a trial unit to a memory is not a comparison.
- Qualify in stages. Samples, then tooling and first articles, then any compliance testing your market requires, then one full production run. Each stage has a pass criterion agreed in advance, and each is a point at which you can stop.
- Inspect the trial run to the same standard as a live one. Same checklist, same AQL sampling, same independent inspection. A trial that gets softer scrutiny than production tells you nothing useful.
- Price the whole thing, not the unit price. Tooling, qualification, testing, freight, the management time, and the yield you lost during ramp. Then compare like for like — which is harder than it sounds, and the reason three quotes are rarely comparable.
- Decide on evidence, and keep both sources alive. If the second source passes, give it enough continuing volume to stay warm. A qualified supplier you never order from is a phone number, not a second source.
What we do
The work, and its honest limits.
Map your actual input origins. We go through your bill of materials with your suppliers and establish where the significant components come from, rather than accepting the assembly address as the answer.
Assess category by category. Whether an alternative exists at your quality and volume — not in principle, but for the specific thing you buy and the standard you hold it to.
Model the real cost. Tooling, qualification, freight, and the quality risk of a new relationship, alongside the option you already have: a second qualified factory inside China. Finding and vetting that alternative is work we do either way.
Then give you a recommendation. Including “stay and second-source within China”, and including “this category should leave.”
And say what we can't do. Our people are in China. We are not going to claim a footprint in a country where we do not have one, and where a category should move outside our coverage we will say so plainly rather than take the project and improvise. That is the same reasoning behind being open about what we are in the first place.
Engagement model
How this is priced.
- Dual-sourcing assessment Free, no obligation, and the result isn't gated.
- Second-source project in China Quoted per project on scope, agreed in writing before work starts.
- Trial run management Quoted per request on scope, agreed in writing before work starts.
- Trial run inspection US$268 per visit — one factory, one visit, photo-documented report within 24 hours.
- Ongoing procurement Monthly retainer based on order volume and supplier count.