Every promise of faster delivery depends on space that cities are simultaneously trying to build over. Same-day and next-day delivery, urban servicing, trade counter access and rapid replenishment for city-centre retail all require logistics capacity close to the customer. Yet the industrial land that makes proximity possible has spent two decades losing a straightforward argument with housing, offices and other higher-value uses for the same sites. Something in that equation eventually has to give, and in London — the UK city where the pressure has built up longest and is best documented — the evidence suggests it already has. For landlords, investors and developers, the more difficult question is not simply whether a city needs industrial land, but which sites can genuinely accommodate the logistics operations it depends upon, and whether the income those operations can support justifies retaining or redeveloping that land.
How much land has actually gone
The scale of the loss is not a matter of impression. Greater London Authority evidence published in 2026 confirms that London lost 18% of its industrial land between 2001 and 2020, including a 9% reduction in land specifically used for warehousing. In Inner London the losses were considerably greater: 37% of industrial land and 39% of warehousing land disappeared over the same period. Those are land-use figures, not a direct measure of warehouse floorspace, and they matter partly because the loss was geographically concentrated in areas where proximity to central customers has particular operational value. The capital’s industrial floorspace vacancy rate, recorded at 3.2% in 2021, reflected pressure in the market at that time, but cannot be treated as a current measure of availability.
The 2026 market is more complicated than the historic shortage narrative alone suggests. Savills’ July 2026 London and South East logistics research reported a combined vacancy rate of 9.48%, with 13.4 million sq ft available across 77 large units. Those figures might appear to undermine the idea that London has insufficient warehousing capacity, but the market being measured consists of much larger logistics buildings, not the smaller, tightly located premises required by every city-serving operation. A distribution business looking for a secure inner-urban depot with space for twenty or thirty vans to load, turn and depart cannot necessarily substitute a vacant 150,000 sq ft warehouse beside a motorway interchange, even if the larger building’s rent per square foot looks attractive.
Nor should scarcity be assumed throughout the smaller-unit market. Savills’ separate May 2026 study of London’s mid-box industrial market recorded 2.77 million sq ft available across 63 units in the 20,000–100,000 sq ft range, equivalent to approximately 3.98 years of supply at the recent rate of take-up. The Savills figures concern different size bands, locations and market definitions, and neither dataset measures how many sites have the combination of yard space, delivery radius, permitted operating hours and vehicle access needed for intensive last-mile distribution. That distinction is commercially important: a sector can have substantial available floorspace while a particular type of logistics accommodation remains difficult to find, just as a protected industrial location can contain buildings that are no longer attractive to the occupiers for whom its designation was intended.
London’s position is the best documented, not a template that should be imposed indiscriminately on Birmingham, Manchester, Leeds or other major cities. Their industrial supply, road networks, housing values and planning policies differ. What is common to growing urban areas is that new homes, businesses and public services generate logistics requirements that must be accommodated somewhere, whether within the city itself or across a wider distribution network. The question of where those functions sit, and at what cost, becomes more difficult as land capable of handling them is lost.
Why the land keeps losing the argument
The mechanism behind the loss is straightforward economics, even if the direction of the land market has become less predictable. Savills’ 2022 analysis showed that London industrial land values had risen by approximately 175% since 2017 to an average of around £8.3 million per acre at that time. That figure is not a current 2026 valuation benchmark, but it illustrates the competing pressures. Strong demand for industrial accommodation can raise rents, investment values and land prices, while the potential value of the same site for residential or mixed-use development may still encourage its owner to seek a different use where planning policy permits. Industrial land can become increasingly valuable without becoming sufficiently valuable to defeat every competing development appraisal.
That relationship is not fixed. In a separate 2022 assessment, Savills suggested that sites representing around 9% of London’s then residential development pipeline, approximately 130,000 proposed homes, could be at risk of remaining in or returning towards industrial use because of changing relative development economics. That was an assessment of potential viability pressure, not evidence that those homes had actually been replaced by warehouses. Its importance lies in the reversal it identified: where residential construction costs, affordable-housing obligations, finance costs and achievable sales values weaken a housing appraisal, continued industrial occupation or industrial redevelopment can become comparatively attractive even on land previously assumed to be destined for residential use.
For an industrial developer, however, the apparent scarcity of land does not remove the basic test of whether a new building can earn enough to justify its cost. A valuable location may support stronger rents because it saves occupiers delivery time and transport expenditure, but the rent an operator can sustainably pay still depends on the savings the site actually produces. A constrained plot with inadequate loading, difficult access or limited electrical capacity may command a significant price because of its location while supporting a building whose operational compromises make it unattractive to the very logistics occupiers supposedly driving demand. A protected industrial designation may limit residential redevelopment without making an expensive replacement warehouse viable, and a theoretical uplift in floorspace achieved through intensification is of limited value unless it can be let on terms that support the capital required.
The investment opportunity is therefore not simply to buy scarce industrial land, but to identify sites whose access, configuration and planning position make them difficult to replace. Such sites may offer a degree of occupational resilience that does not appear in the floorspace figures alone, particularly where the surrounding urban development makes finding equivalent accommodation increasingly difficult. But scarcity is not a substitute for examining the building, its access arrangements, the tenant’s actual requirements and the alternatives available to that tenant.
What proximity to the customer is really buying
None of this land pressure would matter as much if last-mile logistics could simply relocate further out. For some operators it can, and for others the financial penalty is sufficiently great to change their competitive position. Understanding the difference requires separating the various things that closeness to the customer is actually buying.
Delivery radius and driver hours are the most direct constraint. A van-based distribution operation’s daily capacity is limited by how many stops each driver can complete within a shift, and that number can fall materially when travelling between the depot and the service area absorbs more of the working day. Moving from an inner-urban location to an outer one does not simply increase mileage; it may require more vehicles, drivers or shifts to deliver the same number of parcels, maintain the same replenishment schedule or meet the response times promised to customers. The property saving from moving outward can therefore be offset by an increase in operating expenditure, although the size of that trade-off depends on delivery density, routing, service requirements and the cost of alternative accommodation.
Congestion compounds this in ways that are specific to city driving rather than industrial location generally. Journey times in dense urban traffic are less predictable than those on many peripheral routes, making tightly scheduled deliveries harder to operate across longer distances. Clean-air and vehicle restrictions introduce further considerations according to the city concerned, while the growing use of electric vans can make charging infrastructure and grid capacity part of the property decision rather than a fleet matter dealt with separately. An inner-city depot may shorten delivery rounds but prove expensive to electrify, while a more distant site may offer affordable charging and vehicle standing at the cost of longer daily journeys. The cheapest building and the cheapest delivery operation are not necessarily in the same location.
Yard and parking requirements pull in the opposite direction from land scarcity, which is precisely the tension at the centre of the issue. A last-mile operator may require substantial van standing, loading positions, safe vehicle circulation and enough flexibility to cope with concentrated departures and returns, rather than a large internal storage area in isolation. A building can be highly efficient in its use of internal space yet poorly suited to an operation in which vehicles need to enter, load and leave within a short window. That is one reason last-mile facilities cannot simply shrink their land requirement in proportion to their warehouse floorspace as they move into more expensive urban locations: the building may become smaller, but the operational demands of the yard do not necessarily shrink with it.
The same point applies to the diversity of businesses using industrial land. A trade counter, a maintenance contractor, a food distributor, a building-supplies business and a high-volume parcel depot may all value proximity, but for different reasons and with very different physical requirements. A development that meets one occupier’s needs may be unusable by another despite occupying the same planning category. For investors, the significance is that industrial value cannot be assessed entirely through rental evidence and unit size. The replacement-occupier market may be broad for a well-configured building with adaptable loading and access, but considerably narrower for premises where one particular operating model is the only practical use.
The site is not valuable to a logistics occupier simply because its address is close to customers; it is valuable if that proximity can be converted into sufficient operational savings to justify the cost of occupation. For a landlord or developer, the distinction between the two is where much of the real investment risk sits.
Multi-storey logistics: where the UK evidence stands
If land is the binding constraint, building upward is the obvious theoretical response, and it has been demonstrated at remarkable scale elsewhere. ESR’s Higashi Ogishima distribution centre near Tokyo, completed in 2023, provides approximately 349,000 sq m across nine storeys on a site of around 77,700 sq m. It is a substantial logistics building designed around heavy goods handling, with extensive vehicle-servicing infrastructure and floors capable of supporting industrial loading requirements. Its scale demonstrates what can be achieved where the land market, construction economics and occupational demand support the investment. It does not, however, prove that the same configuration could be replicated economically on a constrained London site at the rents UK occupiers are prepared to pay.
The UK is not starting from zero. SEGRO V-Park Grand Union, completed in North West London in April 2024, delivers 134,500 sq ft of industrial accommodation across six storeys on a 1.7-acre site. Two lower levels provide loading and parking, with cargo lifts serving industrial units above. It is a genuine example of completed industrial intensification, with a range of unit sizes aimed at different occupiers and functions, rather than merely a speculative proposal. But it is not equivalent to a high-throughput distribution depot in which large numbers of goods vehicles can directly access multiple operational floors.
The building’s published specification makes the distinction clear. There are five goods lifts, including two with a 6,500kg capacity, alongside eight allocated van spaces and five goods-vehicle spaces. The upper floors offer different clear heights and floor loadings, with units ranging from relatively small workspaces to substantial individual floors. Those details matter more than the headline number of storeys, because they establish the range of uses the building can realistically accommodate. Lift capacity, loading arrangements, circulation time and vehicle standing determine whether the accommodation can serve a particular distribution operation, just as its floorspace and rent determine whether the operator can afford it.
This is not a criticism of the scheme, which demonstrates one way of delivering industrial workspace on land that could not accommodate an equivalent amount of conventional single-storey development. It is a distinction between two different industrial products. A business moving relatively modest quantities of goods, carrying out production or using industrial space for storage and dispatch may benefit considerably from an upper-floor unit served by goods lifts. A large-scale parcel operation dispatching dozens of vehicles within a tightly controlled morning period could have very different requirements, and additional internal floorspace alone cannot establish whether such an operation would be practical.
Multi-storey construction also changes the development appraisal. Upper floors need stronger structures, appropriate loading, goods circulation, additional fire-safety provision and servicing arrangements that consume both building area and capital. Shared loading facilities can improve land efficiency but may create operational constraints when several occupiers need access simultaneously. The increased complexity of construction has to be justified by achievable income, and the occupational market must be sufficiently broad to support the building beyond its first tenants. A scheme that delivers much more floorspace per acre can still achieve a disappointing return if construction expenditure, shared-area costs and the compromises imposed on occupiers exceed the value of the additional accommodation.
The useful conclusion for developers is therefore narrower than either dismissing multi-storey logistics or treating it as the inevitable solution to urban land shortages. Different forms of industrial intensification are appropriate for different occupiers. Building upward can produce commercially useful additional workspace on scarce land, but it does not automatically replace the loading capacity, vehicle circulation or operating freedom of a ground-level logistics site. Any assessment of what industrial capacity has been retained must consider those functions rather than simply comparing the number of square feet before and after redevelopment.
Co-location is more promising for some uses than others
A less dramatic, and in some circumstances more immediately deliverable, response to urban industrial land pressure lies in mixing uses on the same site rather than choosing between them outright. The adopted 2021 London Plan provides for industrial intensification and co-location, including schemes that combine industrial accommodation with housing where the planning and design conditions can be satisfied. In practice, relatively few developments can meet both sets of operational requirements without compromise, because the uses have fundamentally different needs and tolerate very different neighbours.
Light industrial activity, small workshops, trade counters and service-oriented units with relatively limited vehicle movements can, with careful design, sit reasonably comfortably beneath or beside residential development. Heavier logistics operations are more difficult. A depot with HGV movements through the night or a fleet of vans departing in the early morning creates noise, traffic and servicing requirements that can conflict with residential amenity even where the buildings themselves are carefully separated. Acoustic treatments, access arrangements, internalised loading and operating restrictions may reduce those impacts, but can also add cost or limit the activity that the industrial accommodation is capable of supporting. A redevelopment that replaces ground-level logistics space with lighter industrial units and reports an increase in employment floorspace has not necessarily replaced the logistics function that was lost.
The July 2026 draft London Plan recognises this distinction more explicitly. Proposed Policy GLE2 supports industrial intensification, but identifies considerations including sufficient yard space, freight transport connectivity, long-term viability and the relationship with sensitive neighbouring uses. Where an intensification scheme can accommodate only certain lighter industrial uses, the policy proposes that it should demonstrate how that development will help free up capacity to meet other industrial requirements elsewhere. That is a meaningful planning distinction for developers: the additional floorspace may be acceptable as part of a broader strategy, but it should not necessarily be counted as a direct replacement for every industrial operation previously occupying the land.
The draft also takes a more complicated approach to land release than simply protecting existing industrial designations. Proposed Policy PV5 identifies circumstances in which certain designated industrial sites could be considered for housing, while excluding sites required for strategic needs, important freight-access locations and industrial land within the Central Activities Zone or its three-kilometre buffer, among other criteria. The intention is to balance housing growth with the retention or re-provision of industrial capacity, including cooperation between boroughs in the same industrial property market area. In parallel, the draft contemplates identifying additional industrial land, including in specified Green Belt locations subject to the relevant planning process. It is a significant proposed reallocation of where some of London’s industrial functions might be accommodated, not merely a continuation of the previous approach to industrial land release.
Those policies remain proposals, and the adopted 2021 London Plan continues to provide the strategic planning framework while the new plan proceeds through consultation, examination and adoption. The current consultation closes on 15 October 2026. They nevertheless matter to an investor assessing longer-term development options because they suggest a more selective approach to which sites should remain in industrial use, which might be intensified and where some capacity may need to be provided elsewhere before industrial land is released. Such a framework can influence the prospective value of individual sites without guaranteeing that a proposed redevelopment will receive planning permission.
The commercial difficulty is that re-providing industrial capacity elsewhere is not necessarily the same thing as preserving the economic usefulness of the location being lost. A site with direct access to central London’s customers, suitable loading arrangements and the ability to operate through the night may perform a function that a larger warehouse further out cannot replace without increasing transport time and operating costs. Equally, a poorly configured industrial plot close to housing may be a less effective location for intensive logistics than a modern facility on a well-connected freight corridor. The development and planning challenge is to establish which functions genuinely need to remain close to the city and which can be accommodated more efficiently elsewhere, rather than treating every industrial hectare or square metre as interchangeable.
The scale of the demand this land shortage has to serve
The pressure is not simply historical. The 2026 draft London Plan proposes a benchmark of approximately 1.26 million sq m of additional storage-and-distribution floorspace across London between 2024 and 2036, associated with an indicative 270 hectares of land. The longer-term benchmark reaches approximately 2.81 million sq m and 520 hectares by 2050. Those are planning estimates, not completed projects or guaranteed development requirements for individual sites, and they incorporate assumptions about increasing efficiency through better building layouts, higher floors, automation and intensification. Their value is in demonstrating the scale of accommodation that planners are seeking to make provision for while other demands on land continue to grow.
The geographical distribution is revealing. Within the 2036 benchmark, the draft identifies approximately 320,000 sq m of additional storage-and-distribution floorspace for its Central Services Area, associated with an indicative 70 hectares of land. The draft recognises that some central logistics needs may have to be accommodated in other property market areas because of land constraints, while seeking to minimise the additional freight distance that results. For an industrial investor or developer, that is not merely a planning statistic. It demonstrates the difference between identifying sufficient floorspace somewhere within a metropolitan market and providing it in a location where the intended occupiers can operate efficiently.
The relationship between housing and industrial demand is older than the latest planning proposals. A British Property Federation study by Turley, discussed by RICS, identified approximately 69 sq ft of warehouse floorspace for every existing home in England, using 2017 data. It was a national stock relationship, not a fixed quantity of warehousing required every time another home is built, and the ratio differs substantially between areas serving national distribution networks and those predominantly accommodating local logistics. It cannot simply be multiplied by London’s annual housing target to establish how much additional warehouse floorspace must be built inside the capital. The underlying relationship is nevertheless important: residential development creates additional requirements for deliveries, maintenance, servicing, construction supplies and the broader movement of goods, even where the warehouses supporting those functions sit outside the city’s boundaries.
For landlords and investors, the demand forecasts therefore point towards opportunity but do not remove the need for site-specific judgement. Scarcity can support rents where a property offers a combination of access, operating flexibility and genuinely useful accommodation that competitors struggle to replicate. It can equally encourage developers to propose ambitious intensification schemes whose financial success depends on occupiers paying rents that their operations cannot sustain. An industrial designation can protect a use without making an obsolete building worth rebuilding, just as rising last-mile demand does not mean every warehouse located near an urban population will benefit equally. The investment case must still rest on the property’s actual utility and the income it can generate.
What this comes down to
The individual site-selection question — where should a particular business locate its own facility — is a different problem from the one this article has been addressing. This is a structural question about whether cities can retain enough functional industrial capacity to support the logistics demands generated by the people and businesses who live and work in them, and the London evidence shows how difficult that becomes once land values, housing policy and the physical requirements of distribution begin pulling in different directions. The latest market figures demonstrate why a general shortage claim is insufficient; the emerging planning framework illustrates the extent to which industrial supply now requires deliberate coordination; and completed multi-storey developments show that additional accommodation can be produced without necessarily replacing every activity previously carried out at ground level. Whether planning protection, changes in relative land values, co-location or more intensive construction ultimately produces sufficient capacity remains unresolved. What is increasingly difficult to defend is the assumption that urban industrial land is simply a residual category awaiting a higher-value use, or that replacing its measured floorspace necessarily replaces the function the city depends on it to perform.
This article provides general commercial-property and planning-market analysis, not property investment, planning, transport, valuation or legal advice. Industrial property markets, land values, operating requirements and planning policies differ by location and jurisdiction. Developers, investors, landlords and occupiers should obtain appropriate professional advice before making property acquisitions, development decisions or logistics commitments.
