There is a ritual most readers of this magazine will recognise. The parties arrive at eight-thirty. Opening session by ten, if there is one at all. Caucuses through the morning, positions restated over sandwiches, the first real movement somewhere around three in the afternoon. By six, the mediator is carrying numbers down the corridor. By nine, exhaustion has done what reason could not, and a deal in principle is scribbled onto a term sheet in a conference room that smells of cold coffee. Everyone shakes hands in the car park and goes home half-triumphant, half-hollow.
The one-day mediation is the most successful format in the field's history. It settles an extraordinary proportion of the disputes fed into it, and it has earned its dominance honestly: it concentrates minds, imposes a deadline, contains costs, and fits the diaries of busy neutrals and busier lawyers. Nothing in this column should be read as contempt for it. For a very large class of disputes involving two parties, litigation already on foot, a money claim, well-understood positions, and decision-makers in the room, it remains exactly the right tool.
But something has happened to the field while the format stood still. The one-day model has quietly stopped being a format and become a definition. Mediation, in the minds of most users and many practitioners, simply is the day: the booking, the bundles, the marathon, the car park. And that definition is now colliding with the disputes international mediation is being asked to resolve and with the larger role that courts, institutions, and governments are asking mediation to play. The collision deserves more attention than it is getting.
Where the Day Came From
The single-day model was not handed down from first principles. It is an artefact of a particular history: litigation-adjacent mediation, developed largely in common law jurisdictions, dominated by insured claims and commercial money disputes, priced and scheduled around the availability of neutrals drawn from senior legal practice. Its economics are the economics of the court run-up: one day is what parties would pay for, one day is what diaries would yield, and one day is what a dispute already narrowed by pleadings and disclosure genuinely required. The format then hardened into procedure with mediation agreements drafted around the event, institutional rules assuming it, fee structures built on it, training courses rehearsing it, until the day stopped being a choice and became invisible.
Its virtues are real and should be stated fairly. Deadline pressure is a genuine settlement force; decision fatigue, for all its dangers, dissolves positions that months of correspondence could not. Cost certainty matters to parties. Concentration matters to momentum. The day works, and any argument for moving beyond it must respect the reasons it works.
Where the Day Breaks
The strain shows in the disputes that now define international practice. Consider what the format assumes and then consider a contemporary cross-border case against each assumption.
The day assumes the decision-makers are in the room and can commit by evening. International disputes routinely involve state entities, family conglomerates, consortia, and lenders whose approval processes are collective and cannot be compressed into an afternoon: a subject to which this column will return in its own right. Forcing the end-of-day crunch on such parties does not produce settlement; it produces either deadlock or a deal in principle that dies in committee the following month.
The day assumes the dispute is informationally ripe and that both sides know enough to price risk. Complex construction, energy, and technology disputes frequently arrive at mediation with genuine and legitimate information asymmetries: delay analyses not yet exchanged; quantum built on models the other side has never seen. A single day offers no mechanism to cure that. The result is familiar: parties who cannot move because they cannot responsibly evaluate, and a mediation that fails not for want of will but for want of design.
The design of the day assumes two sides. Its very architecture betrays it: two camps of rooms, one corridor, a mediator shuttling between them carrying offer and counteroffer. Multi-party cases involving employer, contractor, subcontractors, insurers, funders, and public bodies often involve settlement geometries that cannot be negotiated simultaneously in one building in one day. The genuinely difficult work in such cases is sequencing: which bilateral understandings must come first, which parties need separate processes, and how partial settlements are structured without stranding anyone. Sequencing is a matter of weeks, not hours.
The day assumes the settlement is an endpoint. Yet many international disputes arise within continuing relationships, including joint ventures, long-term supply, and concession arrangements, where the parties do not need a cheque but a restructured future. Agreements of that kind, drafted at eleven at night by exhausted people, are notoriously fragile. Every experienced mediator has seen the late-night deal in principle unravel at the documentation stage, and most of us have privately wondered how many of our car-park triumphs were quietly renegotiated, resented, or breached within the year. Settlement rates are measured at the door. Durability is not measured at all, and the one-day model is optimized for the metric we count, not the outcome parties need.
There is, finally, an uncomfortable point about one of the day's greatest weapons. Exhaustion is not a neutral solvent. It dissolves the resistance of the tired, the outnumbered, the less well-resourced, and those operating in a second language, faster than it dissolves the resistance of the strong. A process that relies on fatigue to close deals should at least be honest that fatigue is doing some of the work consent is supposed to do.
Why This Matters Now
If this were only a craft observation, it could wait. It cannot because the institutional environment is changing around the format. Courts in England and Wales can now order parties to mediate, following the Court of Appeal's decision in Churchill and the rule changes that followed; other jurisdictions are moving in similar directions, and mandatory or strongly encouraged mediation is spreading through court systems, statutes, and procurement frameworks worldwide. Mediation is being asked to carry structural load: to relieve court backlogs, to sit inside project governance, to manage the waves of disputes generated by geopolitical and supply-chain disruption.
A process conscripted into the justice system on that scale will be judged on more than its settlement rate. It will be judged on the quality of consent it produces, the durability of its outcomes, and its fitness for the disputes assigned to it. If the field's answer to every referred dispute, however complex, however multi-party, however unripe, is the same eight-hour marathon, then mediation will fail a visible fraction of those cases in a very public way, and the failures will be attributed to mediation itself rather than to a format mismatch. The one-day model, in other words, is now a reputational risk for the field, not merely a design choice within it.
Mediation as Process, Not Event
The alternative is not longer days. It is the recovery of an older and truer idea: that mediation is a managed process of which meetings are components, not the thing itself.
In practice, that looks less exotic than it sounds. It begins with a diagnostic phase in which the mediator, appointed early, works with the parties to map the real obstacles to settlement, whether they be informational, structural, relational, or political, before anyone books a conference suite. It continues with designed exchange: not disclosure in the litigation sense, but targeted, proportionate steps that make responsible evaluation possible, agreed and supervised within the mediation. It proceeds through sequenced sessions: some plenary, some bilateral, some involving only principals, some only experts, all arranged around the settlement geometry of the case and the approval rhythms of the parties. It uses interim agreements deliberately: standstills, project-preservation measures, partial settlements, agreed frameworks initialled subject to board confirmation. And it treats implementation as part of the process, with the mediator retained through documentation and, where relationships continue, available for review points thereafter.
None of this abandons the day's virtues. Deadline pressure is designed in rather than relied upon: phases have end dates, sessions have agendas, and the mediator manages momentum as actively as any single-day neutral manages a corridor. Concentrated sessions still happen, but often more productively because they occur when the case is ripe rather than when the diary said so. What changes is that time becomes a design variable rather than a constraint inherited from habit.
The objections are predictable and deserve answers. Cost: a phased process costs more than a day, but much less than the arbitration it replaces or the failed settlement it prevents; the honest comparison is not with the day, but with what follows when the day fails. Momentum: the fear that process kills urgency is real, and the answer is mediator discipline, not format nostalgia, because a drifting extended mediation is a badly run one, exactly as a drifting hearing day is. Appetite: parties, we are told, want the day. Some do. Many, in my experience, want the day because it is the only product on the shelf, and respond with visible relief when offered a process shaped to their dispute instead.
What Would Have to Change
Moving the field beyond the single-day default is not primarily a matter of persuading mediators, most of whom already improvise phased processes in their complex cases, while the formal architecture pretends otherwise. It is a matter of infrastructure. Fee models built on the daily rate penalize exactly the front-end design work that complex cases need; institutions should develop and normalize process-based and phase-based pricing. Rules and model clauses drafted around the event should be redrafted around the process, with express provision for staged mediation, interim agreements, and mediator continuity. Training that rehearses the choreography of the day should teach diagnosis, sequencing, and process architecture as core competences. And courts and legislators integrating mediation into their systems should resist writing the one-day assumption into their schemes, as a referral to mediation should be a referral to a process proportionate to the dispute, not to a date in a diary.
The single day will remain the right answer for many disputes, perhaps most. But a mature profession distinguishes its habits from its principles. The principle was never the day. The principle was self-determination, assisted by a skilled neutral, within a process fitted to the conflict. For the disputes now arriving at international mediation's door, which are typically sprawling, layered, continuing, and increasingly sent to us by systems that expect results, fitting the process to the conflict will more and more often mean letting the mediation take the time the dispute actually requires. The car-park handshake at nine o'clock will always have its place in the field's folklore. It should no longer define its practice.