China's 8.6% retail vacancy: why rent rises alone no longer work
With national commercial vacancy at 8.6% and rent growth under pressure, operators are shifting from pure rent models to operations-led income.

China's commercial property squeeze is not a demand problem. It is a structural one: rent alone can no longer carry the asset, and operations have not yet grown into the second leg. According to Shangce Network's "2026 Must-Haves for Commercial Real Estate Operators," redevelopment projects now account for over 46% of the industry's stock, national commercial vacancy stands at 8.6%, and rent growth remains under pressure. NOI, tenant retention and asset securitisation have become the three hard metrics. For operators, the question has changed: it used to be "did we lease it out?" Now it is "does the maths work after we lease it out?"
At 8.6% vacancy, why is anyone still talking about rent increases?
Because raising rent is not wrong in itself. Relying only on raising rent is.
The same Shangce Network article notes that the industry has fully left behind the incremental dividend of "buy land, make money," and that rent growth remains under pressure. Read that in reverse: there is not much room left in pushing rents higher on their own.
In a Xinhua urban renewal case, Gu Dandan, project lead at Beijing's BOM Xipanli, put it plainly: "A single rent model can hardly support the long-term operation of stock renewal projects." Their move was to shift from "space landlord" to "content and industry service provider," building a composite revenue system of "base rent + turnover share + self-operated business + content services + industry support." The same report gives the numbers: less than 7,000 square metres of actual usable area, driving 2.3 million visits and nearly RMB 200 million in turnover.
Those figures cannot be copied wholesale, but they make one point: rent is an outcome, not a lever. Whether a unit can keep generating turnover is what decides whether you dare negotiate the next round of rent.
Peers are moving operating decisions forward
According to the same Xinhua report, Beijing Datun Power Plant converted idle coal storage warehouses into sports venues. The team started with research on the regional consumption gap, then looked at spatial conditions, and only then fixed the tenant mix. The renovation avoided major demolition and rebuilding; the money went into structural reinforcement, fire safety, water, electricity and HVAC — the parts you cannot see. Project lead Zhang Yang says weekday footfall runs at 600 to 800 visits a day, weekend at 1,000 to 1,200, and revenue in the first three months covered that year's operating costs.
According to the Ministry of Commerce retail innovation casebook released by the China National Commercial Information Centre, Nanjing Xinbai took a full-chain digital approach: consumer big data for user profiling and customer segmentation, smart guides, self-checkout and AI footfall capture deployed together, connecting the online and offline fulfilment chains of more than 400 brands in the mall.
See the pattern? What these two have in common is not the systems they bought. It is that they moved the judgements along the "leasing — signing — rent collection — service" chain from experience to records. Who is being followed up, how far along, when this account should be issued, whether this month's meter reading has been taken — once these actions live only in people's heads and in WeChat, they will leak as soon as you scale.
Where exactly does AI change this chain?
It changes the part where "nobody is watching the things that should be done."
Daily commercial leasing work is largely repetitive but not difficult: a leasing lead with no follow-up for three days, a bill that has come due without a reminder, a meter reading that should have been billed. None of it is hard. What is hard is having someone remember it every day.
SHEYU's AI commercial leasing system (SHEYU AImarket) hands this link to an AI steward: automatic follow-up on leasing leads, reminders on overdue bills, meter reading and billing, contract activation and lease termination. Write operations pass through a two-gate "preview → confirm" flow, with a full audit trail. Underneath, contracts, bills, commissions and meter readings run on a database transaction engine: activation triggers billing, termination triggers settlement. The company's own description is that "market owners don't need to learn the system — if you can talk, you can use it."
The boundary should be stated clearly too: it records accurately, chases on time and reconciles correctly the operating actions that have already happened. It does not decide for you whether a unit should have its rent cut, and it does not guarantee you will fill the space. The tool solves "leakage" and "disorder." It does not solve "what is this location actually worth."
Where should smaller operators start?
Start by getting the books and the collections under control. Do not open with a big renovation.
According to Cushman & Wakefield data cited in a Xinhua Finance report, US shopping mall vacancy in the second quarter of 2026 was just 6.0%, below the historical average of 7.4%, with about 94% of space leased — prime locations in mature markets remain tight. Location still holds value; the question is whether you can reliably collect that value.
For smaller operators, the order is roughly this: first put unit status, contracts, bills and meter readings onto a set of accounts you can actually check; then talk about leasing conversion and composite revenue. Skip the first step and every number in the second is a guess.
AI will not lease an empty unit by itself, but it can stop money you are owed from quietly leaking away.
Whether to adopt a system now depends on whether you hold a dozen units or a hundred. With a dozen units and one person on top of it, a spreadsheet is no embarrassment. Once units multiply, contracts get messy and collection reminders scatter, the human brain starts making mistakes — that is when it is time to change tools.