Singapore's F&B failure rate is widely cited at 60 to 70 percent within three years. Location is the variable most consistently associated with which operators survive and which do not. We mapped rent-to-demand disconnect across 100 subzones to identify the highest-exposure districts.
Source: URA Rental Transactions Q2 2026 · SiteMetriq F&B scoring model · 100 subzones
The core problem
The highest commercial rents in our dataset are not in the highest-demand locations. Several districts with PSF above SGD 17 have F&B demand scores that are among the weakest we measured. The best value district in our dataset costs 58 percent less per sqft than the worst value district, with a break-even daily transaction count less than half as demanding.
Location risk is a function of two things: what it costs to be there, and how much demand the location generates. We express this as a ratio of F&B demand score to monthly occupancy cost. A high score and low break-even is the safe combination. A low score and high break-even is where operators consistently fail.
The findings below are at planning area level. Within each district, individual subzones vary significantly. One subzone in Queenstown scores 74. Another in the same district scores 36. The planning area average does not tell you which unit to take.
High rent relative to the F&B demand signal we measure. Not all subzones within these areas are poor choices.
Among the highest PSF in Singapore with demand scores that do not support the occupancy cost for most F&B concepts.
Commercial space adjacent to industrial zones commands high PSF but daytime population is near zero in several subzones.
High PSF with low transit and demand scores in several subzones outside the main shopping node.
Large HDB stock creates an assumption of demand that the subzone-level data does not fully support.
Lower PSF with stronger demand signals. Subzone-level variation still matters significantly.
Lowest viable commercial PSF in our dataset with strong F&B demand signals in several subzones.
Below-median PSF with some of the highest F&B scores in the dataset.
Significant variation within the planning area. Some subzones score very high. Others do not.
Higher footfall-to-rent ratio than CBD equivalents. Strong evening and weekend demand.
Dense residential catchment. Several subzones score well for GP clinic and daily F&B.
1.Treating residential density as a proxy for demand
Large HDB populations create an assumption of foot traffic. But if that population shops at the town centre two MRT stops away, the local commercial unit does not benefit from the density.
2.Ignoring rent trend direction
A falling market improves your negotiating position. An operator who signed at peak rates in a now-softening district is paying above market. Checking trend direction before signing takes five minutes.
3.Not running the break-even number before viewing
Monthly rent is in the listing. Your average ticket is something you know. The daily transaction requirement to cover rent at a 10 percent rent-to-revenue ratio is calculable in 30 seconds. Most operators do not do it until after they have already decided they like the unit.
District data is the starting point, not the answer
The district-level findings above show where to look and where to be cautious. Whether a specific address stacks up depends on the exact unit. A SiteMetriq report calculates your break-even, benchmarks the quoted rent against URA subzone transactions, and scores the demand at your specific address.
Analyse my address - SGD 59