Alternative Service: When the Front Door Fails

English courts have permitted service through social platforms for well over a decade, an injunction was famously served via Twitter as early as 2009, and orders for service by Facebook, Instagram and LinkedIn messages are now unremarkable where the account is shown to be the defendant's and active.

Those two showings carry the application. Ownership: profile photographs, biographical details, mutual connections, posts referencing known events. Activity: recent posts, comments or logins, evidence the account is a living channel rather than an abandoned page. Screenshots with visible dates, properly exhibited, do the work.

Open-source capture of this kind is a core investigative skill, and it pairs naturally with physical serve attempts: the defendant who will not answer the door but posts daily stories has, helpfully, identified the channel that will reach them. Courts increasingly order service by a combination, message plus email plus first class post, to maximise the chance of actual notice.

The high-water mark of judicial pragmatism came in D'Aloia v Persons Unknown in 2022, where the High Court permitted service of proceedings by non-fungible token, airdropped into the very crypto wallets connected to an alleged fraud, alongside email. The reasoning was pure Abela: it was the method most likely to bring the claim to the attention of defendants who existed, as far as the claimant knew, only as wallet addresses.

Crypto fraud litigation has since made wallet-drop service a recognised tool, and the same flexibility extends elsewhere: courts have ordered service via online contact forms, gaming platforms and messaging apps where the evidence showed that was where the defendant actually lived online.

The lesson generalises. Rule 6.15 is method-agnostic; what it cares about is effectiveness. If your evidence shows a channel the defendant demonstrably uses, the novelty of the channel is no objection.

Rule 6.15 can authorise alternative service on a defendant outside the jurisdiction, but the analysis gains a diplomatic layer. Where service treaties such as the Hague Convention apply, official channels are the default, and courts are slower to bypass them; the authorities require more than mere inconvenience or ordinary delay before treaty routes are sidestepped.

That said, the cases show orders being made where treaty service would take extraordinarily long, where the destination state's channels have broken down, where urgency is genuine, injunctions especially, or where, as in Abela itself, no convention applied at all.

Two practical notes. The court will not generally authorise a method that is positively unlawful where the defendant is. And even for foreign defendants, email and messaging orders remain the workhorses, jurisdiction does not change where people read their messages. Specialist advice on the destination country's position is worth its cost here.

There is a boundary, and the Supreme Court drew it in Cameron v Liverpool Victoria. A hit-and-run victim sued the unidentified driver and sought to proceed by alternative service on the insurer. The court refused: proceedings cannot be validly served, by any method, on a person who is not merely anonymous but untraceable, because no form of service could reasonably be expected to bring the claim to their attention.

The distinction that matters is between defendants who are anonymous but reachable, the pseudonymous wallet-holder, the unknown occupiers of a property, the online troll behind an active account, and those who are simply gone. The first group can be served by whatever channel demonstrably reaches them. The second cannot be sued into the void.

For practitioners the takeaway is a sequencing rule: identification and tracing come before service strategy. A thorough trace either finds the defendant, converts them into the reachable category, or documents the impossibility, which itself informs the client's realistic options.

A lender instructs service of proceedings on a guarantor. The trace confirms his address; three attended visits across a fortnight find a occupied house, a car that moves between visits, and a neighbour who confirms he lives there but works away. On the fourth visit a man matching his description answers, denies being him, and closes the door.

The file now contains dated attempt notes, the trace report, the denial incident, and screenshots showing his LinkedIn account posting that same week plus an email address he used with the lender a month earlier. The application seeks service by email and LinkedIn message, with first class post to the confirmed address as belt and braces, and asks for a deemed service date two business days after sending.

The order is granted on paper. Service is effected the same afternoon, the certificate exhibits the send records and delivery indicators, and when no acknowledgment arrives, default judgment follows without a murmur. Nothing in that sequence was luck; every element was manufactured by process. That is the entire craft of alternative service.