The case law nobody voted for

4 July 2026 · Carl Heaton · Deep dive Commentary

This filing is part of a Steelwise deep dive into 25 years of .uk domain disputes, published as the service that decides them moves to WIPO on 7 July 2026.

No parliament passed it, no judge presides over it, and no law report publishes it. Yet over 25 years, Nominet's Dispute Resolution Service (DRS), the private system that settles .uk domain disputes, grew a genuine body of case law. Of the 5,525 published decisions we analysed, 1,274 cite at least one earlier decision, and together they form a citation network of 3,365 links reaching 928 earlier cases. Because DRS citations use unambiguous case numbers, this part of our analysis is close to perfectly reliable, and it shows a legal system building itself from scratch, one dispute at a time.

Built on losses

Here is the counter-intuitive part: the foundations of this case law are disproportionately the cases the brands lost. A routine transfer of an obvious squat sets no precedent, because there is nothing to argue about. The decisions that get cited for decades are the hard boundary cases, and hard boundary cases are the ones complainants lose or win only on appeal.

The most-cited decision in DRS history is seiko-shop.co.uk from 2002, cited 255 times, which drew the line on when a reseller may use a brand it genuinely sells. Second is verbatim.co.uk from 2007, cited 181 times: the holder of an ordinary English word that happened to be someone's trade mark kept it, and every ordinary-word dispute since has argued in its shadow. Third by raw count is toshiba-laptop-battery.co.uk, a decision that was itself overturned on appeal. Alongside them sit scoobydoo.co.uk (a cultural name, kept by its holder) and parmaham.co.uk (you cannot own a geographic indication).

Raw citation counts also flatter the famous. Rank the network by authority instead, counting citations from decisions that are themselves cited, and a 2001 case called xigris.co.uk surfaces at number three despite only 10 direct citations. It is a precedent of precedents: the early decisions that taught the citing decisions how to reason.

One correction is worth stating plainly, because a naive read of the data gets it wrong. The share of all decisions citing precedent fell across the eras, from 42% to 17%, which looks like a system losing interest in its own law. It is an artefact. The 2008 summary-decision fast lane grew to carry most cases, and summary decisions are short-form documents that almost never cite anything. Among full, reasoned decisions, precedent engagement held steady at 43 to 55% across the entire 25 years.

The self-correcting tier

Appeals are the deepest layer. There were only 86 in 22 years, roughly one and a half in every hundred decisions, but a three-expert appeal panel disagreed with the original expert 40% of the time: 34 of the 86 changed the outcome, correcting in both directions: 20 turned a loss into a transfer, and 14 undid one. The overturn rate has recently been high, 48% since 2018, and the data alone cannot say whether that reflects harder cases being appealed or panels growing more interventionist. Appeal decisions run roughly three-quarters longer than first-instance ones, and the panels' output became precedent in turn: Verbatim, the second-most-cited decision in the whole record, is itself an appeal ruling. The panels built their own house law.

The man who threads it all together

The verbatim.co.uk respondent was Michael Toth. Having won the second-most-cited precedent in DRS history, Toth later lost emirates.co.uk on appeal, and sued. In Toth v Emirates in 2012, the High Court held that DRS decisions are not reviewable in court: the expert system is final. The same man wrote one of the tribunal's most important pages and then proved, the hard way, that its pages are the last word.

That is what makes this body of law remarkable. It binds a meaningful slice of British commercial life, it is final in practice, and it exists entirely because 5,525 sets of parties argued and a few hundred experts wrote careful decisions. On Tuesday its administration moves to WIPO. The policy is unchanged, so the precedents still govern. But the library that holds them is changing address, and that move raises questions of its own.

The series

About the data. The numbers in this filing come from our analysis of every published DRS decision: 5,525 disputes decided between September 2001 and July 2026. Citations are matched on explicit case numbers and are high-fidelity.

If you spot an error in this filing, tell us and we will correct it and note the change.

Built on losses

The self-correcting tier

The man who threads it all together

The series

Further reading