Ho Chi Minh City stopped counting projects and started pricing its land.

The instinct during a boom is to bank everything that clears the door. Vietnam is having one of those years, with registered FDI up 61% in the first half and disbursement running at an 18-year high for the period. So the choice Ho Chi Minh City has made is the interesting one: at exactly the moment the money is easiest to take, the city has decided to take less of it. The logic is not modesty. It is arithmetic. Serviced land inside a functioning metropolis is close to fixed, and every hectare handed to a low-value assembly plant is a hectare that cannot later hold something worth ten times as much. When the constraint is land rather than demand, the metric that matters stops being how much you attracted and becomes how much value each hectare carries.

The city has turned that idea into a rule an intake officer can actually apply. A data centre now has to clear a minimum investment density of more than $100 million per hectare to qualify for space in the high-tech park. High-tech manufacturing has to clear $20.4 million, and research and development $11.5 million. This is the quiet innovation in what otherwise reads as a familiar quality-over-quantity speech. Most agencies say they want better projects and then approve whatever arrives, because quality is a feeling and the pipeline is a number. A density floor converts the feeling into a threshold, printed and non-negotiable, that a project manager can hold up on the first call. It also does the political work of saying no for you: the parcel is not available at that density, and there is no minister to lobby about it.

The wider market says the bet is aimed in the right direction. UNCTAD's latest read has strategic sectors, data centres, AI infrastructure and semiconductors, now accounting for 44% of global greenfield project value, up from 16% at the start of the decade, with semiconductor investment compounding at 54% a year. Greenfield manufacturing outside those sectors, the assembly and light industry that a density floor is designed to filter out, fell 17% over the same stretch. A city that raises its threshold is, in effect, fishing where the fish are moving. The catch is that every other ambitious city has read the same report and set the same targets, so a high floor buys a seat at a crowded table rather than a table of one.

Which is why the more instructive half of the story is on the supply side, where the floor is either made real or exposed as a slogan. Ho Chi Minh City is not only filtering, it is building the ground that a high density needs to be feasible. A newly approved 194.84-hectare expansion in Long Phuoc is being laid out on a green, clean and digital model aimed at carbon neutrality, and a separate 52.92-hectare site is being turned, with Vietnam National University, into a strategic technology centre folding manufacturing, AI and data-centre space into one campus alongside the labs and the engineers who staff them. A density floor without power, serviced land and talent behind it is just a way to turn investors away. Paired with the campuses and the pipeline, it becomes a credible offer: bring the high-value project and the site can actually carry it.

The honest counterweight is that a density floor optimises for capital per hectare, and capital per hectare is not jobs per hectare. The high-value, high-density project the city is now courting typically employs fewer people per dollar than the assembly work it is designed to crowd out, which is a real tension for anywhere whose mandate is employment rather than prestige. Ho Chi Minh City's answer is to pair the floor with a workforce push, enrolling thousands into semiconductor and related fields and courting design and R&D mandates that need local engineers rather than just local floor space. Whether that closes the jobs gap is the open question, and it is the question every agency tempted to copy the floor has to answer before it does, because the trade is not hidden. It is the whole point of the policy.

For a practitioner anywhere, the transferable move is not the specific dollar figure, which is calibrated to one city's land prices and ambitions. It is the act of pricing the land at all. Most agencies never write down what a hectare of their scarce serviced estate is supposed to return, so they end up discovering the answer by accident, one over-generous approval at a time. Ho Chi Minh City wrote the number down first and let it govern the pipeline. That is available to a small EDO with a single industrial park as much as to a megacity, and it costs nothing but the discipline to set the figure and then hold it when a large, low-density deal shows up wanting a discount.