For a hundred years, the person best equipped to read a jury was the lawyer who had tried the most cases. Instinct, sharpened over decades in front of juries, was the only method available. It still has real value. It also has a structural problem: it doesn't scale, it doesn't compound, and it walks out the door the day the lawyer retires or the outside consultant's engagement ends.
Litigation intelligence is the evolution. It is the structured analysis of the people who decide cases, juror segmentation methodology, live juror scoring during voir dire, and verdict modeling. It is also the preparation behind the case the jurors hear: testing trial hypotheses against multiple panel compositions and analyzing before a single juror is ever seated. All of it runs through one trial strategy software system a trial team shares, rather than a single expert's private judgment.
What it is not. It is not Artificial Intelligence (AI) replacing a trial lawyer's judgment, nor is it a black box that spits out a verdict prediction. The attorney directs every matter and makes every strike decision; the system supplements that process with quantifiable insight and applies AI only where it earns its place: processing large volumes of case evidence and running counsel-controlled simulations to pressure-test themes before trial. Scoring itself runs on a structured juror segmentation methodology, not AI pattern-matching but the same discipline that proven marketing segmentation borrows from. The moment a system claims to replace legal judgment, or leans on AI past the point where it's actually more accurate, is the moment no serious trial team should trust it.
The three components of litigation intelligence.
Juror segmentation begins before the panel is seated, and the discipline behind it isn't new. Marketers have used behavioral segmentation for decades because demographics alone no longer predict behavior; a 45-year-old suburban homeowner and a 45-year-old urban renter can look identical on a spreadsheet yet buy nothing alike. Litigation intelligence (also called jury analytics or litigation risk intelligence) asks the same kind of question in a different room: not just who a juror is, but how they tend to decide, their relationship to authority, how they assign responsibility, and how they process risk. Two jurors identical on a demographic sheet can process the same evidence in opposite ways. Because segmentation doesn't require a seated jury, it goes to work early, shaping the case narrative and giving counsel a read on jury response before mediation or settlement talks begin.
Live scoring happens during voir dire strategy, and it's the closest thing litigation has to a focus group, except this one runs under oath: jurors are legally bound to answer truthfully, a data quality advantage marketing research never gets. As jurors respond, the trial team scores them in real time against the themes that matter for that case using jury consulting software. This is also where segmentation gets tested, not just applied: hesitation, contradiction, and bias a questionnaire alone would never surface. Segmentation says where to look. Live scoring says whether you were right. The panel stays ranked and visible throughout selection, not reconstructed from notes afterward.
Verdict modeling extends the same structured approach beyond jury selection and into the trial itself. The model doesn't reset once the panel is seated. It carries forward each juror's segmentation traits and the tells surfaced during voir dire, then layers on how that specific panel is likely to process the evidence and testimony as they're presented at trial, rather than on material that lives in discovery but never reaches the courtroom. A juror flagged as skeptical of institutional authority during voir dire, for instance, is likely to weigh a cross-examination that targets a company's internal process differently than the panel around them. The model doesn't predict how any one juror will vote. It gives the trial team a running, evidence-informed read on the seated panel as a whole, updated as the trial develops rather than reconstructed after a verdict comes in.
None of these three is new in isolation. Demographic jury research has existed for decades. What's new is the chain, with juror segmentation methodology feeding live scoring, which in turn feeds verdict modeling, each stage sharpening the next, rather than three vendors handing off static reports. That compounding, not just sharing a platform, is the differentiator in modern jury selection consulting and trial strategy software.
Why this matters now. Nuclear verdicts, defined as jury awards of $10 million or more, reached a fifteen-year high in 2023. A few plaintiff firms have already adopted structured jury analytics. Most haven't. They're winning on instinct and a built-in sympathy advantage against corporate and insurance defendants, with the numbers already favoring them. Defense counsel, the carriers funding that spend, and the corporate counsel absorbing the verdict are running the same instinct-based process from the other side, propped up by consultants who leave nothing behind when the engagement ends. Whoever moves first, on either side, sets the pace with litigation intelligence.
The part that actually changes the economics. Traditional jury consulting is a one-time purchase. You pay for a consultant's judgment on one matter, and when the engagement ends, that judgment leaves with them. Nothing carries forward to the next case. Litigation intelligence, run through a platform rather than a single outside expert, works differently: every matter adds to a body of retained intelligence the firm keeps. Juror profiles, scoring models, and case themes compound. The tenth matter benefits from the first nine. That compounding is the actual distinction between a consulting expense and an owned asset, and it's the reason firms are starting to ask not 'should we do jury research' but 'should we own the capability or keep renting it one matter at a time.'
That question, VoirPro built to answer it, is covered in more depth on the Voirtex Platform page and the FINDS framework page and in our full FAQ.