Why Distinguishing Owner-Occupiers from Tenants Changes the Risk Calculus

Every commercial insurer knows that the risk profile of a property is not determined by its bricks and mortar alone. The occupant — their legal status, financial resilience, operational behaviour, and tenure — is frequently the dominant variable in claims frequency, business interruption severity, and liability exposure, particularly for the commercial owner occupier. The commercial owner occupier plays a critical role in the overall risk assessment process.

Yet the industry has historically relied on proxies. Business rates data tells you who pays for a property, not who owns it. Electoral registers tell you who is present, not who is liable. Companies House tells you who is registered at an address, not who operates from it, nor how many other entities share that same post box.

Consider the scale of what is missing. The Doorda commercial property dataset covers 2,505,122 properties across England, Scotland, Wales, and Northern Ireland. Within that dataset 312,171 properties have been flagged as owner-occupied — meaning the occupant is the same legal entity as the freehold or leasehold owner. The remainder — 2,192,951 properties (87.50%) — are tenant-occupied. The business trading from those premises is a different legal entity from the property owner. Their connection to the location is contractual, not capital.

In addition, understanding the specific needs of a commercial owner occupier is essential for tailored insurance solutions.

Of the total, 513,728 properties (20.50%) are classified as freehold, 222,182 (8.90%) as leasehold, and 1,769,212 (70.60%) have no recorded owner type — reflecting the structural difficulty of resolving ownership data at scale across the UK’s fragmented land registry.

These proportions are not uniform across sectors. Retail — the largest category with 984,033 properties — shows an owner-occupancy rate of 13.7%. Industrial & Manufacturing properties, despite conventional wisdom suggesting higher rates, come in at 11.9%. Personal Services (hairdressing, beauty salons) lead the sectors at 14.1%, while Healthcare & Education sit at 10.4%. At the lower end, Telecoms & Media properties are just 3.5% owner-occupied, and Transport & Automotive properties sit at 4.9%.

Doorda’s Commercial Property dataset highlights the importance of the commercial owner occupier in evaluating risk dynamics in the market. The intricate relationship between tenants and the commercial owner occupier must be considered to accurately assess potential risks.

This comprehensive dataset is crucial for understanding the unique needs and risks associated with the commercial owner occupier.

The Commercial Property dataset from Doorda that surfaces this distinction — through a dedicated owner-occupier flag (occupant_owner, boolean) — reconciles freehold ownership data, leasehold interests, operational occupancy records, and company registry information into a single, queryable intelligence layer. Every property record carries not only this flag but also the occupant name (covering 1,060,919 unique occupant entities), occupant company number, commercial owner name, commercial owner company number (identifying 226,544 unique commercial owners), and commercial owner type — making it possible to verify, not just assume, whether the business trading from a premises is the same entity that owns it.

For years, underwriters have been making occupancy decisions on incomplete signals. The gap in the market is not more data — it is connected data that resolves a single, surprisingly difficult question:

Is the commercial occupier of this property also its owner?

This article explains why that question matters more than many risk models acknowledge, why it is harder to answer than it sounds, and what becomes possible once you can answer it at scale — with monthly refresh cycles and change detection as the engine of ongoing intelligence.

---

1. The Owner-Occupier Blind Spot in UK Commercial Property

In UK commercial real estate, the distinction between owner-occupier and tenant is not merely a legal technicality — it is a structural differentiator of financial behaviour, operational stability, and risk persistence.

Consider two identical properties on the same industrial estate in the West Midlands. Both are occupied by small manufacturing firms. Both generate similar turnover. Both sit within the same local authority boundary. On paper, their risk profiles look interchangeable.

But one owns the freehold. The other leases month-to-month.

The owner-occupier has made a capital commitment to that location. They have a balance sheet reason to maintain the property, invest in fire safety upgrades, and ensure business continuity after disruption. Their cost base includes mortgage or capital depreciation, not fluctuating rent. They cannot easily relocate — the property is an asset, not an expense line.

The tenant, by contrast, has a fundamentally different incentive structure. Their sunk cost in the location is limited to a deposit and fit-out. If the business model shifts, if the landlord raises rent, or if the local economy softens, relocation is a viable — sometimes rational — option.

This distinction has direct underwriting implications. Owner-occupiers exhibit lower vacancy risk, higher investment in premises, and greater stability in risk exposure over time. Tenants — particularly those on short leases or with low rental commitments — represent a more fluid, harder-to-predict risk segment. Interestingly, the dataset shows that tenant-occupied properties actually exhibit a marginally longer average tenure than owner-occupied ones (7.6 years vs 6.7 years), which may reflect the fact that many owner-occupied freeholds are acquired specifically for business premises and may be sold on change of ownership, whereas tenancies — particularly those held by larger, established businesses — can persist across decades.

Yet most commercial property datasets in the UK do not resolve this question cleanly. The Land Registry records ownership transfers, but does not track operational occupancy. Companies House links legal entities to registered addresses, but not to their actual trading locations, and critically does not distinguish between a head office, a shared serviced office, a dormant shell, or an active trading site. Business rates records identify the ratepayer, but not the freehold owner — and rates avoidance through lease structures and empty property relief claims is widespread across certain sectors.

The Commercial Property dataset underpinning this analysis resolves this gap. It contains a dedicated owner-occupier flag (occupant_owner, boolean) for every commercial property record, derived from a reconciliation of freehold title, leasehold interest, operational occupancy, and company registry data. The result is a national picture of exactly which businesses own the premises they trade from — and which do not.

---

2. The Data Challenge: Why This Is Harder Than It Looks

Building a reliable owner-occupier flag at scale across the UK’s commercial property stock is not a trivial data engineering problem. It requires the reconciliation of at least four distinct data domains, each with its own coverage gaps, update cadences, and identity resolution challenges.

Freehold ownership is recorded by HM Land Registry, but the register is not fully digitised for all titles, ownership information lags behind transactional reality by months or years, and corporate ownership structures — holding companies, subsidiaries, offshore vehicles — obscure the beneficial owner behind layers of legal entities. A property whose freehold is held by “Acme Properties (UK) Limited” may, in practice, be owned by a pension fund, a private equity vehicle, or the operating company next door. This is precisely why 70.60% of properties in the dataset have no recorded owner type — the ownership trail simply cannot always be resolved at the time of data collection.

Occupancy is even less standardised. The dataset captures 1,060,919 unique occupant names across nearly 2 million occupied records. A company registered at an address with Companies House may be the legal occupant, a mail-forwarding service, a shell entity, or a subsidiary sharing premises with a parent group. Business rates records identify the party liable for payment, but lease structures frequently split liability between landlord and tenant in ways that obscure who is actually trading from the premises. On average, 6.5 companies are registered at each address — another signal that many commercial addresses serve as correspondence hubs rather than single-occupancy trading premises.

Leasehold ownership adds another layer. In many commercial properties, the leasehold interest is itself a valuable asset, traded and refinanced independently of the freehold. The dataset captures 222,182 leasehold properties (8.90%) — but the real figure may be significantly higher, given that many leasehold interests are recorded under alternative ownership classifications or are not registered at all.

The commercial owner occupier represents a business commitment that can greatly influence underwriting decisions.

Entity resolution is the hardest problem of all. The same business may appear under different legal names, trading names, registered addresses, and company numbers across different data sources. A pub operator, for example, may hold the freehold under one entity, trade under another, be registered for VAT at a third address, and appear in business rates records under a fourth. Without a robust entity resolution layer — linking occupier names, company numbers, ownership structures, and property identifiers — the owner-occupier question remains unanswerable at scale.

The dataset addresses this through a multi-pass matching process. Each record carries not only the occupant name and occupant company number but also the commercial owner and commercial owner company number (where available), enabling direct comparison of legal identities. Fields such as n_companies_registered_at_address (mean: 6.5), n_registered_appointment_at_address (mean: 11.4), and n_disqualified_directors (6,477 addresses have at least one) provide additional resolution layers to distinguish genuine trading premises from correspondence addresses or corporate shells.

This is why the distinction has remained stubbornly absent from most commercial property intelligence products. It is not that nobody thought of it. It is that the data integration challenge is genuinely difficult — and resource-intensive to maintain at monthly frequency.

---

3. What the Dataset Reveals: A National Picture of Ownership Occupancy

The commercial property dataset covers 2,505,122 properties across virtually every local authority in England, Wales and Scotland, spanning multiple categories of commercial use — retail, industrial, office, leisure, hospitality, and more.

Each property record contains approximately 56 structured fields that span five data domains:

Domain	Key Fields
Occupancy	Occupant name, company number, occupation date, SIC codes 1 & 2, years at address, owner-occupier flag
Ownership	Commercial owner name, owner company number, owner type, occupant_owner flag
Property	Floor count (avg: 1.3), basement present (120,080 properties), size (m², avg: 355 m²), rental value (avg: £37,782), business rates, site internals
Location	Full address, postcode, UPRN/UDPRN, output area, LSOA, local authority, county, lat/long coordinates
Regulatory & Risk	FSA hygiene score, FSA business type, gambling licence type, Companies House compliance (97.1% compliant), disqualified directors

Among the most significant cross-domain correlations is the relationship between ownership status and occupancy tenure. The data reveals:

Occupancy Type	Properties	Avg Years at Address
Tenant-occupied	1,481,347	7.6 years
Owner-occupied	312,088	6.7 years

This includes nearly a decade of continuous observations — allowing temporal analysis of occupancy patterns across economic cycles.

Sector-level owner-occupancy rates reveal a clear hierarchy:

Sector	Total Properties	Owner-Occupied	% Owner-Occupied
Personal Services	22,578	3,186	14.1%
Retail	984,033	135,023	13.7%
Industrial & Manufacturing	5,994	711	11.9%
Healthcare & Education	74,061	7,695	10.4%
Office	461,331	40,434	8.8%
Leisure & Hospitality	183,406	12,444	6.8%
Other	329,589	19,046	5.8%
Transport & Automotive	89,632	4,418	4.9%
Telecoms & Media	74,971	2,616	3.5%
Uncategorised	279,527	86,598	31.0%

The dataset also captures site internals as structured map fields (location_internals_m2 and location_internals_count) — granular breakdowns of floor space allocated to different functions such as kitchen areas, cold storage, server rooms, manufacturing zones, retail floor space, and office accommodation. When cross-referenced with ownership status, these internals reveal distinct patterns: owner-occupied industrial sites, for example, tend to allocate a higher proportion of space to specialised infrastructure (cold storage, heavy manufacturing) than equivalent tenanted properties, reflecting longer investment horizons and site-specific capital expenditure.

The average property in the dataset has a rental value of £37,782, an average size of 355 m², and an average cost per m² of £328. The occupant SIC code fields (primary and secondary) enable sector-level segmentation of ownership patterns — and the sector-level variation above demonstrates that ownership status is far from uniform.

For a commercial insurer, this sector-level ownership intelligence translates directly into portfolio segmentation capability. A book of retail properties — the largest sector with 984,033 properties — can be segmented by ownership status into those 135,023 owner-occupied locations with capital commitment versus the 848,145 tenant-occupied locations with contractual tenure.

---

4. Practical Applications in Commercial Insurance Underwriting

For insurers, distinguishing between a commercial owner occupier and a tenant is vital to understanding risk profiles.

When the owner-occupier signal is resolved across a national commercial property database, a range of practical underwriting applications become operational.

Risk persistence modelling. Owner-occupiers have made a capital allocation decision that typically implies a multi-year commitment. The dataset captures this directly through n_years_occupant_address, a computed field reflecting how long the current occupant has been associated with that property. The data shows an average tenure of 6.7 years for owner-occupiers and 7.6 years for tenants — suggesting that tenant stability in many sectors may be higher than assumed, but also that the reasons for departure differ materially between the two groups. For insurers writing property damage, business interruption, or liability cover on commercial portfolios, the ability to segment by ownership status provides a structural lens on risk stability that renewal-based data alone cannot.

Vacancy risk scoring. Vacancy is one of the largest sources of uncertainty in commercial property insurance. A property that becomes vacant shifts from a known risk (operational business with active management) to an unknown one (empty building with unmonitored systems, increased vandalism risk, potential squatters). With 120,080 properties having basements — physical spaces particularly vulnerable to deterioration when unoccupied — the consequences of unremediated vacancy are significant. The monthly refresh cycle of this dataset means these signals can be detected within weeks, not quarters.

SME behavioural segmentation. Small and medium enterprises represent a large share of commercial insurance premium, but their risk profiles are notoriously heterogeneous. The owner-occupier flag provides a clean, observable segmentation variable. An SME that owns its premises is likely to exhibit different financial behaviour — lower leverage on the property, longer planning horizons, higher investment in compliance and maintenance — than an equivalent SME operating from leased space. Fields such as companies_house_compliance (97.1% compliant nationally), n_disqualified_directors (6,477 addresses have at least one), and n_companies_registered_at_address (mean: 6.5 per address) add further resolution. An owner-occupied property with a compliant company, no disqualified directors, and a single registered entity at the address is a meaningfully different risk from a multi-occupancy tenant address with 11.4 registered appointments and compliance flags — even if their SIC codes and turnovers are identical.

Portfolio exposure aggregation. For insurers writing large commercial portfolios, understanding the concentration of ownership structures across the book provides an additional dimension of exposure management. A portfolio heavily weighted toward tenant-occupied properties in Inner London (where owner-occupancy is just 8.9% across 184,967 properties) faces a different risk profile to one tilted toward Outer London (where the rate rises to 14.0% across 128,472 properties). The geographic variation is significant: Greater Manchester sits at 8.6%, Kent at 11.8%, and Essex at 11.1%.

Business continuity assessment. After a major incident — flood, fire, infrastructure failure — the resilience of the insured business depends partly on their ability to return to the same premises. Owner-occupiers have stronger incentives and fewer barriers to rebuilding in place. Tenants, particularly those with lease breaks approaching, may simply relocate, triggering business interruption claims that extend beyond the physical reinstatement period. Ownership status is a leading indicator of post-incident behavioural response.

---

5. The Strategic Value of Monthly Refresh and Change Detection

A static owner-occupier flag is useful. A dynamic one — updated monthly with change detection — is transformative.

The dataset is refreshed on a monthly cycle, meaning every property record carries the version of these attributes as they stood at the point of the latest refresh. This temporal granularity enables insurers to:

Recognising the characteristics of a commercial owner occupier helps in formulating effective risk management strategies.

1. Detect transitions — When a property changes from owner-occupied to tenant-occupied (or vice versa), it is captured in the next refresh cycle. A sale-and-leaseback event, for example, flips the occupant_owner flag from true to false and simultaneously updates the commercial_owner and occupation_date fields. The change signal itself — the delta between two monthly snapshots — becomes a data point with underwriting significance.
2. Monitor occupancy tenure — The n_years_occupant_address field increments with each refresh, providing a running measure of occupancy stability. A sudden drop in this value (indicating a new occupant) is a trigger for re-assessment.
3. Track ownership structure changes — Changes to the commercial_owner_name, commercial_owner_number, or commercial_owner_type fields may signal corporate restructuring, acquisition, or distress — all of which carry implications for the underlying risk.
4. Observe sector-level trends — By aggregating monthly snapshots, insurers can build trend data showing which sectors are seeing rising or falling rates of owner-occupancy, providing early warning of structural shifts in their portfolio exposure.

When a previously owner-occupied property transitions to tenant occupation, the signal is often a precursor to broader structural change. The owner may have sold and leased back to release capital, indicating financial pressure. The business may have been acquired by a group that consolidates property ownership centrally. The freehold may have been purchased by an institutional investor, changing the landlord-tenant dynamic. Each of these transitions carries underwriting implications detectable at the next monthly refresh.

Similarly, the detection of new owner-occupiers entering the dataset — a business buying its premises for the first time — is a signal of growth, confidence, and capital investment. These are precisely the businesses that may be underserved by their current insurance arrangements and receptive to a more sophisticated risk assessment.

Change detection transforms the owner-occupier signal from a static feature into a leading indicator of business behaviour.

---

6. Building an Intelligence Layer: The Occupier-Owner-Property Triangle

The owner-occupier flag is most valuable not in isolation, but as part of a connected intelligence layer that links three dimensions of commercial property risk.

The occupier dimension. Who is trading from this property? The dataset captures 1,060,919 unique occupant entities across nearly 2 million occupied records. What is their legal structure, their industry classification (SIC code 1 and SIC code 2), their Companies House compliance history (national rate: 97.1%), their director track record (6,477 addresses have disqualified directors)? Do they operate multiple sites, and if so, what is the ownership status of each? How many other companies are registered at this address (mean: 6.5), and how many appointments (11.4 on average)?

The owner dimension. Who holds the freehold? The dataset identifies 226,544 unique commercial owners. Who holds the leasehold interest? Are they the same entity as the occupier (occupant_owner = true), a parent company, a third-party investor, or an offshore vehicle? What is their portfolio of properties, and what does that portfolio reveal about their investment strategy and risk appetite? The commercial_owner_type field provides classification — with 513,728 freehold and 222,182 leasehold interests identified.

The property dimension. What are the physical characteristics of the site — floor area (avg: 355 m²), number of floors (avg: 1.3), basement presence (120,080 properties with basements), site internals broken down by function (location_internals_m2)? What is the rental value (avg: £37,782), the business rates liability, the estimated turnover? The average cost per m² rent stands at £328. How do these characteristics correlate with ownership status?

When these three dimensions are linked through a consistent entity resolution framework — connecting occupier names, company numbers, ownership structures, and property identifiers — patterns emerge that are invisible in any single dataset. A portfolio of properties that appear individually low-risk may, when analysed together, reveal concentrated exposure to a single owner with a deteriorating financial profile. A set of occupiers that appear stable may, when cross-referenced with ownership data, reveal a pattern of short-term leaseholds in properties owned by distressed landlords.

Additional intelligence layers enrich this picture further. The FSA hygiene score and FSA business type fields provide regulatory intelligence for food-related premises. The gambling licence type field identifies properties with gaming exposure. The seats field provides capacity data for hospitality and leisure venues. The location_internals_m2 and location_internals_count fields break down internal space allocation across functional categories — from cold storage to server rooms to kitchen areas — enabling property-level granularity unmatched in conventional datasets.

This is not merely more data. It is a differentiated intelligence layer that enables risk assessment at a level of precision that the current generation of commercial property data products cannot deliver.

---

The Future of Occupancy Intelligence

The commercial insurance industry is entering a period in which data differentiation is becoming a competitive necessity rather than a nice-to-have. Rate adequacy is under pressure from alternative capital. Distribution channels are fragmenting. Regulatory expectations around risk modelling are rising.

In this environment, the ability to answer a single question — does the occupier own this property? — with accuracy, timeliness, and breadth of coverage is a structural advantage.

The dataset exists. 2,505,122 properties. 312,171 owner-occupied and 2,192,951 tenant-occupied. Over a million unique occupant entities reconciled against 226,544 unique commercial owners. 56 structured fields across five data domains, refreshed monthly. The technology to reconcile freehold ownership, leasehold interests, occupancy, and entity resolution at scale is operational.

What is missing is the willingness to treat occupancy intelligence as a core underwriting input rather than an enrichment layer applied after the fact.

For insurers, managing agents, and brokers who recognise that the distinction between owner-occupier and tenant is one of the most informative signals in commercial property risk, the opportunity is clear: build it into your models, your portfolio analytics, and your engagement strategy now — before it becomes standard practice across the market.

By focusing on the commercial owner occupier, insurers can better tailor their offerings to meet specific business needs.

---

We are actively working with commercial insurers to integrate this intelligence layer into underwriting workflows, portfolio analytics, and risk models. If the distinction between owner-occupier and tenant is relevant to your book — and it almost certainly is — we would welcome a conversation about pilot programmes, data partnerships, or collaborative research.

The dataset covers 2,505,122 commercial properties across the UK, refreshed monthly, spanning 56 fields across occupancy, ownership, property characteristics, company records, and regulatory intelligence. We are happy to tailor extracts, provide sample data matched to your portfolio geography, or discuss how the owner-occupier flag integrates with existing risk models.

The dynamics of the commercial owner occupier are crucial for understanding market trends and potential risks.

Frequently Asked Question about Commercial Owner Occupier