The information is public. Assembling it is manual, slow, and mostly incomplete — which is itself a finding about transparency.
Establishing which company owns a given British gambling brand is possible. Every element is in the public record: licence registers, corporate filings, terms and conditions.
Assembling those elements into a map of the whole market is a different proposition. It requires joining datasets that were never designed to be joined, resolving corporate names that change through restructuring, and re-checking continuously because acquisitions never stop.
The consequence is that no complete public map exists, and the partial ones that do are the product of slow manual research.
Why the join is hard
The regulator’s register records licensees, not corporate groups. Two licences held by two differently-named subsidiaries of one parent appear as unrelated entries, with nothing linking them.
Bridging that gap means going to Companies House and following ownership chains upward — a process that is workable for one company and laborious for hundreds. Holding structures are often several layers deep, and intermediate entities frequently carry names that resemble neither the brand nor the ultimate parent.
Then the white-label problem. Where a brand runs on a third party’s licence, the register shows the licensee and not the commercial operator, so the corporate trail can lead somewhere that has little to do with who actually runs the site.
What partial maps show
Independent mapping projects have produced group profiles covering the larger operators, and the shape that emerges is consistent: a small number of groups accounting for a large share of the visible brands, with portfolio sizes varying enormously.
One published index, at https://sistersiteshub.co.uk/networks/, illustrates both the pattern and its limits. It carries completed profiles for four operator groups alongside a further seven identified as significant but not yet profiled — a ratio that is a fair reflection of the state of public knowledge generally. The groups awaiting coverage include several of the largest names in the British market.
The portfolio sizes in the completed profiles range from three brands to ten, which is itself informative: “operator group” describes structures of very different scale, and the sister-site relationship means something different at each end of that range.
| Data source | Provides | Missing |
| Gambling Commission register | Licensee, status, trading names | Parent company links |
| Companies House | Ownership chains, directors | Which brands map to which entity |
| Site footers | Operating company per brand | Group context |
| Terms and conditions | Evidence of shared platform | Formal ownership proof |
Why the maps go stale
Even a complete map would begin decaying immediately, and understanding why explains the state of the field better than any single gap.
Ownership changes constantly. Acquisitions complete, brands are sold between groups, licences are surrendered and new ones issued. A profile accurate when written can be wrong within months without anything about it having been sloppy.
Corporate restructuring compounds it. Groups reorganise which subsidiary holds which licence for tax, regulatory or operational reasons, and those changes are recorded in filings rather than announced. A brand can move between legal entities inside the same group with no visible change to customers.
And brands are retired quietly. A name disappearing from a licence’s trading-name list is often the only public trace, and nobody is notified.
The consequence for anyone reading ownership research is that dates matter more than completeness. A profile carrying a recent review date is worth considerably more than a larger, undated one — and the absence of a date is a reasonable signal about how much weight to put on it.
What the gaps mean
The absence of a complete map is not a scandal — no regulator has an obligation to publish one, and the licensing register does what it was designed to do.
But it does mean the burden falls on consumers and independent researchers to assemble a picture from sources that require expertise to combine. Anyone wanting to know whether two sites are the same business faces fifteen minutes of cross-referencing, and most people reasonably decline.
The information asymmetry is the point. Operators know their own group structure precisely. Customers can find it out, with effort, one brand at a time. In a market where the same company frequently runs several apparently competing products, that gap is doing real work.
The reasonable ask
A single field in the licence register recording the ultimate parent would close most of it, and would be trivial to maintain relative to what licensees already report.
Until something like that exists, the practical method remains manual: footer, register, Companies House, then compare terms and conditions for verbatim overlap. It is not elegant, but it works, and it answers the question the marketing is designed not to.
It is worth noting what this is not. None of the difficulty described here implies concealment or wrongdoing by anyone. Every element is disclosed, in the place the law requires it to be disclosed. The problem is purely one of assembly — several honest datasets that were never designed to be read together, and no obligation on anyone to do the joining. That is a solvable problem, and a cheap one to solve.
In the meantime, the reasonable expectation of any research in this area is that it will be partial and dated rather than comprehensive and current. Treat a group profile as a snapshot, check when it was last reviewed, and verify anything that matters against the footer of the site itself. The primary source is always one scroll away, and it is the only version guaranteed to be describing today.
Free confidential support: BeGambleAware. Gambling is for over-18s only.