It was back in the middle of the pandemic that Pocket Living and Lichfields Planning banded together to try and figure out why small sites and SME developers were not making a more material contribution to housing supply. For years, policymakers had talked about the untapped potential of small sites — urban plots capable of delivering between 10 and 150 homes — as a way to unlock housing supply more quickly.
At the time, the idea was to undertake a more longitudinal approach. Studying smaller sites as they progressed through the planning and development cycles over a number of years. Yet the research stumbled into something much more concerning. The system was breaking down in London. Roughly three years into the Mayoral flagship policy of fast track and late stage review testing, the results exposed a system mired in conflict and attrition.
The research and the results
The 2020 research analysed 60 small-site planning permissions in London, representing 2,666 homes, including 485 affordable units. These 60 sites were chosen at random in a lottery system devised by the research team and then a full review of their planning journey was undertaken by qualified planners using a check list assessment of the main issues and their journey time. I am not aware of this kind of more granular research having been conducted using a live data set.
The headline finding was stark. The average determination period was around 60 weeks, compared with the statutory 13-week target and the government’s 26-week planning guarantee. Only one scheme met the 13-week target, and just two met the 26-week benchmark. Around one in five applications took more than two years to secure consent.
At the heart of these delays sat viability testing. The signature policy of the current Mayor’s first term, fast track and review testing was brought in to embed affordable housing into land values for any scheme over 10 units. By 2020, the new regime was in full swing and the research revealed it had released a wave of clip board wielding assessors on London’s unprepared SME sector.
It is all about viability
In three-quarters of cases the research reviewed, viability testing and affordable housing negotiations were identified as the principal barrier. Far from being a light-touch sense-check, viability assessments on small sites were often as detailed, contested and iterative as those applied to major developments, despite the far more limited scope for value engineering or tenure flexibility.
Many schemes were required to submit multiple viability appraisals. Evidence was being revised repeatedly as assumptions on build costs, values and benchmark land value were challenged and reworked. In summary, the research found that over a third of schemes required two or more planning applications before permission was granted, frequently because viability positions were revisited late in the process.
More concerning still was that the research team was finding that SME developers were no longer able to price land effectively on small and medium sized sites over 10 homes. This raised concerns about future supply and the high degree of risk that smaller sites would be resisted by the viability team at the LPA.
Why this mattered
For SMEs and smaller developers, these challengers were not a tolerable inefficiency but an existential risk. Delay translates directly into finance costs, cashflow strain and heightened delivery risk.
Critically, the 2020 research showed that the public value extracted through the small-sites planning process was often modest relative to the time and complexity involved. Affordable housing contributions were frequently limited, yet achieved only after prolonged negotiation. In effect, the system was imposing high transaction costs without reliably securing proportionate public benefit.
This fragility is reflected in more recent national evidence. Across England, permissions on very small sites have fallen sharply over the past decade. In 2024, fours years on from the London based research, just 17,000 homes were approved on small sites of 3 to 9 units, down from an average of 35,000, with the proportion of total planning` permissions granted on sites of 150 units or fewer, plummeting from nearly 20% in 2008 to 6-8% today,
Data and response
One of the original ambitions of the 2020 Lichfields research was not simply to diagnose these problems, but to track them over time. That ambition has quietly been undermined by changes in London’s planning data infrastructure.
At the time, the Pocket team presented the findings to the GLA and warned that without change, the SME sector was under major threat. Permission-level analysis was possible using a GLA-maintained dataset — the former London Development Database — which allowed individual schemes to be tracked with a high degree of granularity. It was hoped that working with the GLA, the framework could be altered to find a more streamlined route for small and medium sites.
Unfortunately, that dataset was superseded by the Planning London Datahub shortly after the research was circulated. Whether this was just bad timing or something else we will never know but the data was lost.
The practical consequence is that repeating the 2020 exercise now requires third-party commercial datasets such as Molior, Glenigan or Landstack. This raises the barrier to independent scrutiny and weakens the ability of researchers, practitioners and policymakers to assess whether small-site policy is actually delivering.
For London, this is a particular problem. The capital’s housing challenge is not just one of scale, but of composition. Small sites, incremental delivery and SME builders are repeatedly cited as part of the solution. Yet without transparent, consistent and publicly accessible data, London risks flying blind on precisely the segment of the market the Government claims to want to unlock.
Small sites will not solve the housing crisis alone. But until the system is redesigned to fit their realities, and until London can properly measure what is happening to them over time, they will remain a missed opportunity hiding in plain sight.
