Who really owns UK e-commerce? What Companies House says about 220,000 online retailers

If you bought something last week from a UK-looking website you had never heard of, there is a reasonable chance the company behind it did not exist a year ago, is owned in full by one person you have never heard of, is already flagged for removal from the Companies House register, or shares a registered address with dozens of other companies that will not exist next year either.

That is not a rhetorical flourish. It is what the live Companies House register looks like when you pull the active universe of companies registered under the SIC code UK online retailers use.

Most shoppers assume that a UK Limited company sitting behind a shop is a real, ongoing UK business. That assumption used to be broadly safe. This report is about why it is no longer safe, and what a shopper can actually work out from the public record before pressing Buy.

How big is the universe of ‘UK online retailers’ on paper?

SIC code 47910, ‘Retail sale via mail order houses or via the Internet’, is the primary code used by UK Limited companies whose business is selling to consumers online. Most legitimate UK e-commerce brands sit under this code, alongside a much larger mass of drop-shipping micro-brands, TikTok shops, and shell companies.

At the time of pulling, the active universe was approximately 215,000 to 220,000 companies. That is not the total number ever registered under this code — dissolved and struck-off entries push the historical total substantially higher. That is only the live, still-active list.

That number sounds like a healthy trading population. It is not. Once you slice it by incorporation date, most of it collapses.

Almost none of them are more than two years old

Working through the register in incorporation-date order, the age distribution is extraordinarily front-loaded. The overwhelming majority of the ‘active UK online retailer’ population was incorporated in the last 24 months. The tail of long-standing UK online businesses is far thinner than the headline count suggests.

In our stratified sample of 220 active companies, the incorporation dates broke down roughly as follows. Only a small minority — a single-digit percentage — were incorporated before 2020. Companies from 2020 to 2022 form a modest layer, mostly built during the pandemic e-commerce boom, many of which have since been struck off but the surviving ones read as more established trading businesses. The vast bulk of the sample — well over half — was incorporated in 2023, 2024 or 2025.

The practical implication is that when a shopper lands on a UK Limited-branded shop they have never seen before, the base rate is that the company is less than 24 months old. The base rate for a similarly unknown high-street sole trader would be very different. On paper the UK online-retail sector looks huge, but demographically it is a sector of infants.

How many are being struck off within their first year?

The Companies House status ‘Active — Proposal to Strike off’ is important context here. It means Companies House has begun the process of removing the company from the register, typically for failing to file confirmation statements or accounts, and the company has not yet objected or resolved the issue. It is a live indicator that a company is functionally abandoned, even if the storefront is still up.

In our sample, the proportion of companies in this state is not spread evenly across incorporation cohorts. It concentrates ferociously in recent ones.

Read the middle row again. Roughly half of the SIC 47910 companies incorporated in November 2024 were already on their way off the register 13 months later. That is not a natural business-failure rate. That is the signature of a formation pipeline in which the incorporation itself is close to disposable.

The pattern is consistent with the way many drop-shipping and short-run marketplaces operate. Register a Limited company, get onto Amazon or Etsy or a payment processor that requires a UK entity, trade for a season, allow the confirmation statement to lapse, and let the company be struck off rather than dissolve it deliberately. The Limited company is not the business. It is a permit slip.

Are the beneficial owners even in the UK?

The person with significant control (PSC) register is the single best public tool for working out who actually owns a UK Limited company. A PSC is any individual (or entity) that holds more than 25% of shares, more than 25% of voting rights, or the right to appoint or remove a majority of directors. Every UK Limited must declare at least one, or file a statement explaining why it cannot.

In our sample of 60 companies with full PSC data pulled, 100% had at least one PSC on record. Not a single one used the ‘no PSC identified’ statement. That is a genuinely good news finding: at the register level, ownership is being declared. The story is in what the declarations say.

The single most striking pattern is where the declared PSCs actually live.

That figure is the visible layer. It counts only PSCs who filed their country of residence honestly and outside the UK. A shopper who never expected to check the register at all would be surprised by the number. A shopper who assumed a ‘UK Ltd’ shop was UK-based in the meaningful sense — with owners, operations and consumer-protection accountability inside the UK — would be more surprised still.

The wider nationality picture is less about foreignness as a concept and more about how much of the register no longer maps onto the mental model of a UK-domiciled business. Across the 60 companies, at least one declared PSC held a non-British nationality in around 43% of cases. That is not a bad thing on its own. It reflects the reality of who is starting UK online businesses in 2024 and 2025. But it does bear on how a shopper should interpret ‘registered in the UK’ as a trust signal.

One person, one company, one owner

Beyond location, the most consistent finding in the PSC data is control concentration. In our sample, 85% of companies had a single PSC — one individual with 75% to 100% of the shares, votes and director-appointment rights.

That is not surprising for a micro-business. A one-owner online shop is a perfectly ordinary structure. What matters is what it tells the shopper. There is no board. There is no partner keeping the operator honest. There is one person, and behind the storefront the entire trading identity of the business rests on their willingness to keep going, keep filing, and keep answering the phone.

A small subset of that pool is younger than the shopper might expect. In our sample, roughly 8% of companies were controlled entirely by a PSC under the age of 25. That is not a judgement about capability — plenty of successful UK businesses are run by people in their early twenties — but it is context. When a Limited company is one person, and that person is a 20-year-old who incorporated three months ago, the ‘Ltd’ in the footer is doing more work than it deserves to.

The formation-agent postcodes

The other pattern that jumps out of the register is address concentration. Formation agents — accountancy firms and virtual-office providers who register hundreds of companies at their own address — are visible and, in themselves, unremarkable. Where it becomes interesting is the density.

In the 220 companies we sampled, a small number of postcodes appeared over and over. The same Cardiff postcode (CF14 8LH) turned up nine times across three cohorts. A West Midlands postcode (CV2 4AQ) turned up six times, exclusively hosting recently-incorporated companies with Chinese-language directors. A cluster of central-London postcodes — EC1V 2NX, EC2A 4NE, WC2H 9JQ, W1W 5PF — each hosted between three and eight companies in the sample. These are not head offices. They are registered-office addresses being used by dozens or hundreds of companies each.

A shared registered address is not, on its own, a red flag. It is standard practice for a small business to use its accountant’s address. It becomes a flag when combined with everything else: a young company, a single PSC, a country of residence outside the UK, an ‘Active — Proposal to Strike off’ status, and a registered address that also hosts thirty other companies in the same state. The combination is what changes the picture. Individually, each field is defensible. Stacked, they describe a class of company that is very unlike the trading businesses most shoppers assume they are dealing with.

What a real, established UK e-commerce business looks like on the register

For contrast, the 2020 to 2022 cohort in our sample reads very differently. Nearly every company had a PSC resident in the UK. Foreign residence was rare. Multi-PSC ownership (husband-and-wife, business partners) appeared at similar rates but felt more consistent with real trading. Strike-off rates were far lower. Registered addresses were more likely to be a real trading premises, a home address, or a genuinely small accountancy practice rather than a hundred-company virtual-office block.

This is roughly the shape a UK shopper’s mental model was built around: a company that has been trading for a few years, owned by identifiable people who live in the UK, with a real address behind it. The signal any given ‘Ltd’ carries in 2026 depends heavily on whether the specific company behind it looks like this older cohort or like the November 2024 cohort. On the current register, more of them look like the latter.

What can a UK shopper actually see for themselves?

The Companies House public register is fully searchable, free, and quick. A shopper can look up any UK Limited company by name and see, in under a minute, the four fields that carry most of the load.

1. Incorporation date. Anything under two years old warrants extra care. Anything under six months old is effectively an unknown quantity.
2. Company status. ‘Active — Proposal to Strike off’ is a genuine red flag. It means the company is functionally abandoned in the eyes of Companies House.
3. PSC country of residence. If the sole controlling PSC does not live in the UK, the ‘registered in the UK’ label is doing very little consumer-protection work. Any dispute will be much harder to escalate.
4. Registered office postcode. If it maps to a virtual-office building that hosts hundreds of companies, that is context. Not a disqualification, but context.

The fields above are all public. The register is not a private database. But the behavioural research on trust checking is clear that shoppers almost never open it. In what UK shoppers actually check before buying we set out that under 4% of shoppers ever scroll to a site’s footer, let alone look up the company. This is why the burden of the trust check falls almost entirely on visible on-site signals, and on whether the site tells the shopper enough to make the check unnecessary.

Why the ‘Ltd’ in the footer is a weaker signal than it used to be

A decade ago, a UK company number and registered office on a checkout page were a strong proxy for ‘this is a real UK business’. Registration friction was low, but the trading population under SIC 47910 was much smaller, and the base rate of ‘UK Ltd on a shop = real UK trading company’ was high. That base rate has moved.

This does not mean UK Limited status is now a bad signal. It means it is not a self-sufficient one. On its own, ‘UK Limited’ is now compatible with a company under 12 months old, controlled entirely by a single PSC outside the UK, registered at a virtual-office address shared with dozens of others, and already on the path to being struck off. The signal earns its weight when the specific company behind the number is not any of those things. Making that visible to a shopper is not something the register can do on its own. It has to be surfaced by the merchant, in a form the shopper does not have to click through to see.

That is a large part of what a verification signal like a UK trust seal is doing. It is not replacing the register. It is compressing the four checks above into a single visible signal that a shopper who was never going to look up Companies House can still see. This is why third-party verification beats trust copy a merchant writes about themselves: the register is authoritative but invisible; footer copy is visible but self-authored; a verification seal is both visible and third-party.

The bottom line

The register is doing what the register does. It is holding the data honestly. The problem is not with Companies House. It is with the gap between what a UK shopper assumes ‘UK Limited’ conveys and what, on the current data, it actually conveys.

In 2026, ‘this is a UK Ltd’ is a necessary condition for consumer-protection accountability, not a sufficient one. Under it sits a much smaller, older, quieter population of genuinely established UK e-commerce companies, and a much larger population of very young, single-owner, often-foreign-resident, often-already-abandoning trading vehicles. The two groups look identical to a shopper who never checks. Any signal that can separate them, at the moment of purchase, is doing work the register cannot do alone.

The straight answer to the question in the title, based on the current data, is this. UK online retail on paper is dominated by companies incorporated in the last two years, controlled overwhelmingly by a single individual, with a meaningful minority of controlling owners resident overseas, and a strike-off rate that reaches nearly half of one-year-old companies in some cohorts. That is who really owns UK e-commerce. The names most shoppers think of — the older, established brands — are a much smaller share of the active register than the raw ‘UK Ltd’ count implies.