China vs Vietnam for manufacturing: an honest comparison.

Disclosure first: skeelx manufactures in China. That's where our process runs. This comparison is still worth writing honestly, because the right answer for your product is the only recommendation worth anything, and for some products the right answer isn't ours to sell.

Where Vietnam genuinely wins

Vietnam has built real depth in labour-intensive categories: garments and footwear above all, furniture and timber goods, and growing electronics assembly capacity as global brands diversify final assembly there. For products in those lanes, especially at volumes where labour content dominates cost, Vietnamese quotes can be genuinely competitive, and tariff arrangements with Australia (AANZFTA and RCEP membership) keep the duty story workable. If your product is sewn, upholstered or assembled more than it is toolmade, Vietnam deserves a place on your quote list.

Where China stays hard to beat

Everything that depends on the ecosystem: tooling-heavy products, electronics with real component depth, products touching many processes, and programs needing fast iteration between design and line. Vietnam's supply chains still import a large share of components and tooling from China, which means "made in Vietnam" often contains a Chinese supply chain with extra freight inside it. For custom-engineered products, the toolmaker density argument alone usually settles it.

The honest decision frame

Ask three questions. Is your category one of Vietnam's proven lanes, or are you asking a young ecosystem to do a mature ecosystem's job? Does your cost structure lean labour or tooling-and-components? And does your risk posture actually require country diversification at your scale, or is that a board's anxiety wearing a supply-chain costume? Quote both where the category is genuinely contested; the landed-cost model settles arguments the headlines can't.

Then look past the first tier. Customers and auditors are starting to ask where the components and materials came from, not just where the carton was packed, and the EU's digital product passports, being phased in product group by product group, push the same way. A Vietnamese assembly line fed by Chinese components puts two countries in your supplier records, and both need documenting. The cheaper quote is only cheaper if you can stand behind what's in it.

The agent-native lens: comparing countries below the headline

When the buyer sends an assistant. "China or Vietnam?" is exactly the question an AI assistant can answer quickly and badly, by summarising whatever diversification coverage it finds. Ask it a category question instead. Name the processes the product needs, the components it uses and the volumes you'll order, then ask where each part would actually come from under each option, traced back past the assembly country. Have it treat any preferential duty rate as unconfirmed until a customs broker has checked the agreement's rules of origin, which turn on how and where the goods were made, not where they were packed. Factories in either country that say plainly which processes happen under their own roof and which are bought in give an assistant something real to weigh.

When the business runs on agents. A product company sourcing from more than one country can have agents keep a supplier map that goes below the first tier: who makes each part, where, and on what evidence. They can hold current quotes from both countries for products in the contested middle, so a switch is costed before it's urgent, and watch for duty and trade-rule changes for a person to assess with the broker. The same map is the first draft of the traceability record described above. Moving a product between countries stays a decision for people, made against the landed-cost model and a risk posture someone has written down; the agents make sure it never starts from a blank page.