The standard mediation runs a familiar shape. Counsel exchange position statements. The mediator opens a joint session and asks each side to summarize. The mediator listens. There is a caucus. There is another caucus. Eventually, somewhere between the third and the seventh exchange of demands, a settlement number emerges. Or it doesn't, and the parties go home.
For a meaningful share of commercial, employment, and construction matters, this shape is fine. The dispute is small enough, the documents are clear enough, and the parties' counsel are sophisticated enough that the mediator's job is essentially to facilitate the negotiation that the parties were already prepared to have.
For the matters that don't fit that description, there is a different shape. We call it record-first mediation, and at Legion it is how every engagement runs.
What "record-first" actually means
Record-first mediation begins before the joint session. After intake and after the parties have agreed on scope, the mediator works through the relevant document set: the operative contract, the amendments, the correspondence that bears on disputed terms, the financial or operational data that the parties' positions depend on, and any expert work that's already been produced.
The point is not to render an opinion. The mediator is a neutral, not a judge. The point is that by the time the joint session opens, the mediator can speak to the documents in detail. When a party characterizes a contract clause, the mediator already knows the clause. When a party references a deposition, the mediator has read it. When a party asserts that the other side delayed, the mediator has the schedule analysis in mind.
The effect on the negotiation is significant. Parties stop arguing about what the documents say, because the documents have already been read by someone who is neutral and informed. The conversation shifts to what the documents mean, and from there to what the parties actually need.
Where the time goes
A standard mediation tends to spend 40 to 60 percent of its time on the threshold question: what happened. The remaining time goes to the harder question: what to do about it. Record-first mediation reverses those proportions.
For matters where the documents are organized and the parties' counsel have done the underlying litigation work, our pre-mediation review takes between four and twelve hours of mediator time, depending on the complexity. That work happens before the joint session opens, on the mediator's clock, and at no separate fee. By the time the parties sit down, the threshold work is done.
The trade-off is that the upfront commitment is larger. We can't run record-first mediation on a same-day notice. We typically need two to four weeks between intake and the joint session, depending on the document volume. For matters where the parties want to mediate next Tuesday, we are not the right choice.
The matters where it produces the largest gains
Record-first mediation produces the most visible gains in three categories of matters.
Document-heavy commercial disputes. Partnership dissolutions, post-M&A earnout fights, joint venture exits, complex contract interpretation matters. These disputes are won and lost on the document set, and a mediator who has worked it is a faster mediator.
Construction matters. Construction is the practice area where record-first is closest to mandatory. The contract, the schedule analyses, the change orders, the daily reports, the RFIs, and the inspection reports together form a record that the parties' positions are built on. A mediator who hasn't read them is at the mercy of whichever party characterizes them better.
Catastrophic personal injury and complex insurance matters. The medical records, the coverage forms, the reservation of rights letters, and the deposition transcripts all bear on the negotiation. Reading them in advance is what allows the mediator to understand what's actually settling and what isn't.
What it doesn't fix
Record-first mediation does not solve the cases where the parties simply disagree on value. If both sides have priced the risk and the gap is too wide for any settlement to bridge, no amount of mediator preparation closes it.
It also does not substitute for litigation discovery in matters where the parties haven't done the discovery yet. We are not a discovery process. We work from the record the parties bring, and if that record is incomplete, the mediation produces an incomplete result.
What it does is move the disputes that are settleable from the joint-session-as-discovery model to the joint-session-as-negotiation model. For most of our practice, that's enough.
What outside counsel should expect
If you are referring a matter to a record-first mediator, expect the following from intake forward.
An early conversation about scope. We will ask what documents are relevant, who has them, what's privileged, and what the parties have already exchanged. We are not asking to relitigate the case. We are asking to understand what we need to read.
A document request that is narrower than discovery and broader than a position statement. We typically ask for the operative contract or contracts, the relevant correspondence, any expert reports, and the parties' best statements of the disputed facts. We are happy to work from the parties' joint stipulations or chronology when those exist.
A pre-mediation caucus that is substantive. Once we have read the materials, we caucus separately with each party's counsel to surface positions, interests, and the levers that haven't yet been pulled. These caucuses are confidential and privileged. They are also where most of the actual movement happens.
A joint session that is structured and shorter than counsel may expect. With the threshold work done, we typically open the joint session with a brief framing, move quickly to the substantive disputes, and run shuttle caucuses as needed. Most matters that close at Legion close in a single joint session.
The investigative discipline behind it
Legion was founded on the premise that mediation should be held to the same evidentiary standard as investigation. Two decades of investigative work taught us that the parties' narratives diverge in predictable places, and that closing the gap requires reading the same documents both sides are reading. The work itself isn't glamorous. It is, however, the work.
If you are evaluating mediators for a complex matter and you want to understand whether record-first is right for the case, we are happy to have that conversation in a 30-minute consultation. The shape of the matter usually tells us within ten minutes whether the framework is going to add value.