In the year to mid-2025, Vietnam shut one door on its shopping internet and quietly opened another. It suspended Temu and Shein, the cross-border marketplaces that flew cheap parcels in from Chinese warehouses, and it stripped the tax break that had made those parcels cheap. Into the space they left, TikTok Shop pointed its feed at goods made and registered at home, building a channel that runs from a livestream to a Vietnamese sewing floor instead of a customs broker. The familiar reading of this is social commerce swallowing retail. What it more exactly shows is a country with one of the world’s largest garment workforces assembling, for the first time, a home market worth sewing for.

The cross-border channel did not fade on its own; policy closed it. Temu and Shein were ordered to suspend Vietnam operations in late 2024 after missing the government’s deadline to register as local operators, as VietnamNet Global, via the South China Morning Post, reported. Vietnam also scrapped the VAT exemption on imported parcels worth under one million dong, roughly $39. The low-value parcel is the unit on which Shein and Temu economics rest, and within months it lost both its price edge and its clean path through customs. That is the door closing.

TikTok Shop spent the same months making home-made supply easy for the feed to find and cheap for sellers to push. By its own account — announced at a May 2026 CSR event and reported through VietnamPlus — it had attached identification tags to more than 200,000 goods from OCOP and other Vietnamese producers. OCOP is the state’s rural-maker certification scheme; the tags carry machine-readable provenance that TikTok Shop says improves domestic-goods visibility in the feed. Provenance is worth something only if it pays, and here it did: sellers in the platform’s weekly Vietnamese-goods slot saw average revenue climb to 3.5 times their 2024 level. The tag converts a patriotic slogan into a visibility signal, and the visibility signal into orders. This is the supply side of the same door, opening.

Call the result a content-to-manufacturer pipeline, because the feed now does two things a marketplace listing never did. It discovers demand in public, since a livestream shows what sells before the order is placed, and it routes that demand to a registered domestic seller who can also be the maker. The demand is not small: TikTok Shop held about 42% of Vietnam’s e-commerce by value in H1 2025, while the larger incumbent, Shopee, grew revenue at 4% (Vietnam Briefing, drawing on Fint Global and Metric data). A shortening loop sits under those numbers, from a video, to a comment thread, to a cutting table that can read the same signal the creator just did. The marketplace sold whatever was in the warehouse; the feed can commission what the warehouse should hold next.

For apparel, this aims a machine at home that was only ever pointed abroad. Vietnam shipped nearly $44 billion of textiles and clothing in 2024 and has overtaken China as the largest garment supplier to the United States, yet close to 70% of its garment firms work on cut-make-trim terms, where a foreign buyer sends the fabric and the design and the local factory only sews (B-Company). The country owns the assembly step of other companies’ brands and almost none of the demand behind them. A domestic livestream channel is the first buyer of scale those factories can reach directly, under labels they own. That is a different business from stitching someone else’s.

The question stops being whether Vietnam can make the clothes, which it plainly can, and becomes whether the feed can make it worth the country’s while to make them for itself.


The thesis has one clean failure case, and it deserves its strongest form. If the garments moving through the feed are finished imports that a locally registered seller merely relabels, the pipeline is customs arbitrage under a national flag, not production. Even the honest version runs into a wall upstream: Vietnam buys in between 64 and 80% of the fabric it sews, some $26 billion worth in 2024, most of it from China (B-Company). A livestream can localise the brand, the design and the stitching; it cannot localise the loom. The cloth still lands on a container from the country whose parcels Vietnam just turned away.

What breaks the arbitrage reading is that policy raised the price of exactly that trick. Ending the low-value exemption and mandating local registration together raised the cost of the imported-parcel model; both platforms were ordered to halt operations before their registrations cleared. The platform’s provenance tags and fee breaks, meanwhile, pay a seller to be a maker instead of a middleman. None of this guarantees that any given dress was sewn in Ho Chi Minh City, but it moves the marginal garment toward home, where eighteen months earlier every incentive pushed it the other way. The fabric point stands, and it should: what is being rebuilt is the front of the garment chain, brand and demand and assembly, on a back end still imported.

If the domestic channel keeps growing while the fabric stays foreign, Vietnam lands on a faster version of the position it already holds. It will own the camera, the cart and the cutting line, and rent the loom from the same place it just blocked at the border. Should TikTok Shop’s home-goods tilt hold and the government carry its protectionist logic upstream into yarn and weaving, the feed becomes the demand engine for an integrated garment industry rather than a shop window bolted to an assembly one. Fail to move upstream, and it stays a better shopfront wired to the same imported cloth. Vietnam has decided who sells to its people; it has not yet decided who supplies the people who sell.