Ask where the growth is in UK PBSA and London still comes up first. The demand depth is genuine, and so is the structural undersupply. The answer is becoming less useful, though, because the question underneath it has changed.
For a decade, getting the city right did most of the work. Yield compression rewarded exposure, and a well-chosen market forgave a great deal at asset level. That era has closed. Value creation is increasingly found in operator-led NOI growth, which puts occupancy management, cost control and CapEx execution at the centre of the return. The demand backdrop has tightened at the same time. The Home Office granted 383,455 sponsored study visas in the year to June 2026, 11% fewer than the previous year and 41% below the 2023 peak.
City selection has not stopped mattering. It has stopped being sufficient, and the evidence for the difference now sits at room level.
The Demand Picture Has Changed Faster Than the Underwriting
Softer demand does not arrive evenly, which is why a city-level view of it now underprices risk. It concentrates in particular cities, institutions and price points, and it separates assets that were genuinely performing from assets that were well positioned in a rising market. UCAS end-of-cycle data shows 31% of UK 18-year-old accepted applicants intended to live at home in 2025, a record high and up from 22% a decade ago. The Office for Students forecasts 42.7% of English providers running a deficit in 2025-26, up from 35.8% in 2024-25. That lands unevenly across a portfolio, not across a market.
What Market-Level Data Cannot Tell You
Provision rates, rent benchmarks and pipeline depth describe competitive intensity at city level with real precision.
What they cannot describe is how a particular building behaves. Two assets in the same submarket, at the same rent, on the same heating system, can deliver entirely different resident experiences and entirely different operating costs. That variance sits below the resolution of market data.
It is the resolution The Utopi Platform is built at. Every reading comes from our own Multisensors™: Proprietary first-party data across 90,000 connected rooms in 13 countries and 63 billion room-level data points, several heating seasons deep. It cannot be modelled from utility bills or EPCs, because at this resolution the readings do not exist anywhere else.
What Does This Look Like in the Real Market?
Sheffield is a reasonable place to test that, precisely because the market-level picture leaves so little room.
Cushman & Wakefield put Sheffield’s student demand pool down by more than 17% between 2022/23 and 2024/25, and the steepest PBSA rent fall in the country at 5.5%. Where competing supply deepens into a demand base that is no longer expanding, rental growth and occupancy are both constrained.
We took room-level readings from four live Sheffield assets on The Utopi Platform, around 2,300 rooms, and tracked indoor temperature and measured occupancy through 2026. In July, with outdoor temperatures at 19°C across the city, two of the assets has almost identical occupancy, 29% and 28% of rooms in use. The share of rooms above 24°C was 48% at one and 80% at the other. Same city, same month, same outdoor conditions, same heating type, comparable occupancy, and a 32 percentage point gap.
The variance holds across all four. In August, rooms above 24°C ranged from 34% to 75%, and average indoor temperature spanned 3.3°C, from 22.7°C to 26.0°C.
In August the least-occupied asset in the set, at 17% of rooms in use, still had 54% of its rooms above 24°C, while an asset at 19% presence sat at 34%.
Occupancy does not explain it: The asset with the lowest presence was not the coolest. Overheating is a characteristic of how a building actually runs, and it varies enormously between assets that look interchangeable on paper.
Where This Leaves Asset-Level Underwriting
The market data question was which cities. The performance question is which assets, and on what evidence. Provision rates will tell you how hard a market is. They will not tell you whether the building you own is running well inside it, and in a cycle where the return comes from operations rather than repricing, that is the gap that matters.
To see what room-level performance data looks like across a live portfolio, book a walkthrough at marketing@utopi.co.uk