Nearly two decades of manufacturing transitions have taught us something simple: moving production is the easy part. Moving the ecosystem that supports it is an entirely different challenge.

I first ran into this in 2009, on a China-to-Vietnam transition for manufacturing kitchen towels. A simple product, yet Vietnam lacked the infrastructure that China handled as routine. Hand-printing wasn’t available, so we switched to machine printing — an acceptable result, but a step below China’s hand-printed finish. Greige cloth, hangtags, wash labels, even the cartons — all of it was shipped in from elsewhere. Vietnam did the final packing and little else.

We call this the “invisible seam”: the space between where production happens and where the capability behind it actually resides.

Fifteen years later, the gap has narrowed dramatically. But it closed gradually and unevenly, and for much of that time, the capability on the floor wasn’t really Vietnamese. Across three distinct waves of relocation — cost-driven, then capacity-driven, then the post-pandemic scramble for “diversification” and geopolitical decoupling — the same failure pattern kept surfacing. The countries, products and reasons changed. The failure pattern did not.

Upholstery fabric made the seam visible. A major Chinese supplier opened a stocking warehouse in Vietnam to serve the factories that had moved there, seeking material closer to the line, and shorter lead times. But a warehouse can only hold so much. It carried a narrow band of high-volume fabrics, while designers kept asking for the newest collections and broader palettes, most of which still sat in China. The shelves barely moved, and within a couple of years, the warehouse closed. You can move inventory closer to a factory; you cannot move the place where new things are developed. Innovation doesn’t run on a warehouse schedule.

Baby furniture, in 2014, showed how staged the progress really was. The factories could handle most of the build locally, but the drawer slides, hinges and special-process paint finishes still came from Chinese suppliers, and every failed finish inspection had to be resolved across the border. The floor itself was still run by a Taiwanese team. The address was Vietnamese; much of the capability was not.

By 2024, that had changed. On my last visits, Vietnamese factories were reading engineering drawings, solving production problems on their own, and dealing with overseas customers directly in English, with no foreign management in between. And yet I was still chasing recurring color variation on finished sofas: the same cushion, across two runs, coming back in noticeably different shades. The cause sat upstream: fabric development, dyeing, and shade control were all still in China. Vietnam still imports the majority of the fabric its furniture industry uses, and that dependency hasn’t changed in fifteen years. It’s structural. The defect surfaced at the end of the line in Vietnam; its root cause never left China.

Three very different products — towels, cribs and sofas — and underneath each, the same structure. Vietnam’s factories have transformed in 15 years. The dependencies 

For years, I asked the wrong question: Who is my factory?

Today, I ask: Who supplies my factory?

Relocation changes the address on the purchase order. It doesn’t move the fabric mill that owns the shade standard, or the engineering team that makes the last-minute calls. If the original country’s operations paused for a single month, which of your “diversified” product lines would actually keep running?

The true test of diversification is whether the capabilities, materials and development processes behind it have truly decoupled. Factories move. Dependencies migrate slowly. Capabilities evolve more slowly still — and some may never move at all.

Effie Qin Qian is a global sourcing and product development professional.