A complaint lands on the desk of the general counsel at Veridian Manufacturing, a mid-sized industrial company. A former distributor alleges breach of a long-term supply contract and demands $4 million. The instinct, under the press of everything else on the docket, is to forward the thing to outside counsel, file an answer, and wait to see what discovery turns up. That instinct is the beginning of a familiar and expensive mistake.

Here is the uncomfortable truth about civil litigation: the most consequential decision in most cases is made in the first month, by whoever decides how hard to think before reacting. A well-executed case assessment is the single most valuable investment a litigator makes at the outset of a dispute. It determines whether a case settles early or grinds through years of discovery. It shapes the budget the client will approve—or revolt against. It identifies the witnesses and documents that will win or lose the case at trial. And it supplies the analytical framework through which every later decision—whether to move to dismiss, whether to mediate, whether to take a settlement offer to the board—gets made. Yet case assessment is also one of the most underinvested activities in corporate litigation, because the pressure to respond to the immediate demand crowds out the deliberate analysis that should precede it. The result is reactive lawyering: decisions made under pressure without the foundation that would have informed them. Cases that should have settled early run through expensive discovery instead; cases that should have been fought settle prematurely because nobody did the work to reveal their strength.

This article is built to prevent that outcome, and it follows Veridian through the whole arc of assessment—from the fact-finding that should begin the moment a dispute surfaces, through the merits analysis that separates a claim that sounds good from one that can actually be proven, to the quantitative valuation that produces a defensible settlement range, and finally to the practical chokepoints—jurisdiction, limitations, and collectibility—that quietly decide a startling number of cases before anyone reaches the merits at all. Along the way it addresses the developments reshaping how litigators assess cases: the fee structures and ethical rules that govern the engagement, the AI-powered analytics that forecast judicial behavior, the technology-assisted review that has slashed e-discovery costs, and the third-party funding that is rewriting the economics of who sues and who pays.

Working backward from the goal

Effective assessment begins not with the complaint but with the client's goal, because the two are not the same thing. The complaint defines the legal dispute; the client's goal defines the strategic objective. Veridian, facing the distributor's suit, may care less about the abstract contract question than about protecting an ongoing relationship with other distributors who are watching, avoiding a precedent that invites copycat claims, and resolving the matter before it eats its small management team alive. A technology company facing a patent assertion may care less about the patent's actual merits than about the signal a settlement sends to the next assertion entity in line. Understanding the client's true objectives lets counsel calibrate every part of the assessment to what actually matters—quantifying risk, finding the cheapest path to the client's preferred outcome, identifying early-exit opportunities, and reducing exposure to ongoing cost and liability. The skilled litigator works backward from the destination, not forward from the pleading.

That orientation also dictates who is in the room. Assessment is not a solo act of legal brilliance; it is a structured conversation among trial counsel, the client's business people who actually know the facts, and—where the stakes warrant—a damages expert and a coverage lawyer brought in early rather than as an afterthought. The goal at the outset is not an answer. It is the right list of questions, asked in the right order. Those questions, in roughly the order a disciplined assessor takes them up, are the rest of this article.

The golden hour: pre-litigation assessment

The most cost-effective window for resolving a dispute opens before a complaint is filed—before discovery obligations attach and before the parties entrench. The work should begin the moment litigation becomes foreseeable, not when the complaint arrives, because the advantages of early action cannot be recreated later at the same cost. Witness memories are freshest in the weeks after the events; documents that exist today may be overwritten tomorrow; and the client's appetite for early settlement is highest before the sunk cost of litigation makes settlement feel like surrender. In complex, high-exposure matters, counsel should complete the foundational steps within 30 to 60 days of learning of a potential claim.

Often there is a built-in opportunity to exploit. Many jurisdictions and contracts require pre-suit notice—30 days or more—before a lawsuit can be filed, and some consumer-protection statutes, like Massachusetts's Chapter 93A, include pre-suit demand requirements that hand the defendant a chance to respond before litigation begins. That notice period is strategic time: it lets counsel investigate the factual basis, probe the legal theories, find the weaknesses in the claimant's position, and decide whether early resolution is wise. Sometimes the investigation reveals that the client can cure the alleged violation within the notice window, eliminating or shrinking exposure; sometimes it reveals that the claim is meritless and a firm early response—or a well-pitched demand letter of the client's own—will discourage the plaintiff from proceeding at all. (On crafting that response, see our guide to writing a demand letter.)

The litigation hold

As soon as counsel anticipates litigation—again, not when a complaint is filed—the duty to preserve evidence attaches, and a litigation hold must follow. The duty arises when litigation is "reasonably anticipated," meaning the point at which a reasonable party would foresee it. The consequences of getting this wrong are governed by Federal Rule of Civil Procedure 37(e): if electronically stored information that should have been preserved is lost because a party failed to take reasonable steps, and it cannot be restored, a court may order measures "no greater than necessary to cure the prejudice"; and if the court finds the party "acted with the intent to deprive another party of the information's use," it may impose the severe sanctions of an adverse-inference instruction, a presumption that the lost information was unfavorable, or even dismissal or default. A defensible hold issues to everyone who may possess relevant information (including former employees and third parties under the client's control); it specifically identifies the information and subject matters to preserve; it actually suspends routine document destruction; it is monitored for compliance rather than merely issued; and it is updated as the dispute evolves. For Veridian, the hold must reach not just email but the messaging platforms, shared drives, and ERP systems where the supply-contract records live—because a hold that misses where the data actually resides is little better than no hold at all.

Fact-finding and the proof matrix

Once preservation is secured, the assessment turns to structured fact-finding. The investigation should proceed from the periphery inward—documents and context first, then supporting witnesses, then the key decision-makers—so that the most important interviews are conducted by a well-prepared questioner who already knows what the documents say. Witnesses should be interviewed individually, not in groups, because group settings let recollections quietly converge into a tidy but less accurate consensus. Every critical fact should be corroborated by at least one document, because facts resting on memory alone are fragile: memories fade, witnesses leave, and uncorroborated testimony is easily impeached. The investigation should be organized around the elements of the claims and defenses, identifying for each the documents that help, the documents that hurt, and—most dangerously—the documents that should exist but do not, because opposing counsel will ask why an expected record is missing, and an unexplained absence can be more damaging than an unfavorable document that can at least be explained.

Collection is only half the job; verification is the other half. Counsel must cross-check facts across sources, compare documents against recollections, and stress-test the internal consistency of the client's narrative before the opposing party does. A useful technique is the red-team exercise: assign a team member to build the best possible case for the other side using the same facts. This routinely surfaces vulnerabilities that confirmation bias would otherwise hide—and if your own team can poke holes in the narrative, opposing counsel will find those holes and others.

From this work emerges the proof matrix, one of the most effective tools at this stage: a structured document mapping each element of every claim and defense to the specific witnesses and documents that will prove or disprove it, together with anticipated evidentiary hurdles and a candid strong/weak/uncertain rating for each element. The matrix forces the team to think concretely about proof rather than theory, exposes the gaps that discovery must fill, and provides the framework for evaluating settlement offers against the realistic likelihood of proving each element at trial. A complementary discipline is to draft, as a thought exercise, high-level opening and closing statements for both sides—if counsel cannot articulate a compelling closing for the client, the assessment needs revisiting before a dollar more is spent.

The merits: can the plaintiff prove every element?

Here is where most superficial assessments go wrong. They ask whether the client "has a good case" as if that were a single, holistic judgment. It is not. A civil claim is not a vibe; it is a list of elements, each of which the plaintiff must prove, and the omission of any one is fatal no matter how sympathetic the rest of the story. The disciplined assessor takes each cause of action apart element by element, applies the actual governing law to the actual facts, and rates the client's position on each element separately. A breach-of-contract claim, in most jurisdictions, requires the existence of a valid contract, the plaintiff's own performance or excuse for non-performance, the defendant's breach, and resulting damages. Veridian's distributor may have an airtight contract and an obvious breach—and still lose if it cannot prove it performed its own obligations, or cannot tie the breach to a quantifiable loss. The matrix earns its keep precisely here, because it prevents counsel from being seduced by the strong elements into ignoring the weak one that decides the case.

Burden of proof: who must prove what, and how convincingly

Two questions ride on top of the elements. First, who carries the burden of proof? Generally the party asserting a claim or affirmative defense bears it, but the allocation is not always intuitive—affirmative defenses (statute of limitations, accord and satisfaction, failure to mitigate) shift the burden to the defendant, and a few claims carry burden-shifting frameworks that move the load back and forth. Second, and just as important, what is the standard of proof? The default in civil litigation is "preponderance of the evidence"—more likely than not, the proverbial fifty-percent-plus-a-feather. But a meaningful set of claims demands "clear and convincing evidence," a materially higher bar that sits between preponderance and the criminal "beyond a reasonable doubt." Fraud, in many jurisdictions; punitive-damages findings, in many others; the validity of a patent, which an accused infringer must prove invalid by clear and convincing evidence under Microsoft Corp. v. i4i Ltd. Partnership, 564 U.S. 91 (2011)—all of these raise the bar, and a case that is a coin-flip under preponderance may be a clear loser under clear-and-convincing. An assessment that ignores the standard of proof is assessing the wrong question.

Liability versus damages: the two-headed valuation

Perhaps the most common analytical error in case valuation is collapsing two distinct questions into one. Liability asks whether the defendant is legally responsible. Damages asks how much that responsibility is worth. They are independent variables, and a case can be strong on one and weak on the other in any combination. A plaintiff with a 90% chance of proving liability but a damages theory that, realistically tested, supports $200,000 rather than the $4 million pleaded has a far less valuable case than the pleaded number suggests. Conversely, a defendant facing only a 30% chance of being found liable but a catastrophic damages exposure if it loses cannot treat the case as a minor nuisance; 30% of a company-ending number is still a company-threatening risk.

The disciplined move is to assess liability and damages separately, assign each its own probability and range, and combine them deliberately rather than intuitively. For damages, that means understanding the governing measure—expectation damages in contract (the benefit of the bargain), reliance or restitution where expectation cannot be proven, compensatory and possibly punitive damages in tort, and the specialized regimes that govern intellectual property, discussed below. It means scrutinizing causation (are these damages actually traceable to the breach, or to market forces the defendant did not cause?), foreseeability (the venerable rule of Hadley v. Baxendale still does real work in contract cases), the duty to mitigate, and any contractual limitations—liquidated-damages clauses, consequential-damages waivers, and caps—that can convert a frightening number into a manageable one. For Veridian, a well-drafted limitation-of-liability clause in the supply contract may be worth more than every factual defense combined.

The legal review, done honestly

Before the proof matrix can be trusted, the legal review must be done without flinching. Counsel must work through the elements of each cause of action applied to the specific facts, the statutes of limitations and any tolling arguments (more on these shortly), the available affirmative defenses—jurisdictional, immunity, contractual limitations like forum-selection and arbitration clauses, and the ordinary factual defenses—the damages theories and the realistic range of recovery, and, often overlooked, the applicable jury instructions and verdict forms, which define the questions the jury will actually answer and are therefore the real rules of the game. Reviewing them early grounds the assessment in how the case will be tried rather than how it looks in the abstract. The resulting client briefing should be frank rather than optimistic: the client is far better served by an honest account of the weaknesses in its position than by reassurance, and counsel should present options rather than conclusions, because the client—not counsel—decides whether to settle, mediate, arbitrate, or litigate, and cannot decide rationally without the full picture. For Veridian's general counsel, a briefing that candidly rates the breach claim as stronger than the company hoped is worth far more than one that tells the board what it wants to hear and collapses under the weight of discovery six months later.

Three threshold questions that decide cases before the merits

Long before a jury weighs a single element, a civil case can die—or be born into the wrong forum, or be worthless even if won—on three threshold questions that inexperienced assessors treat as afterthoughts. They are not afterthoughts. They are gatekeepers, and a rigorous assessment runs all three first, because there is no point modeling trial outcomes for a claim that is time-barred, in a court that has no power over the defendant, against a party who could never pay.

1. Personal jurisdiction and venue: does this court have power, and is this the right one?

A court cannot enter a binding judgment against a defendant over whom it lacks personal jurisdiction, and the constitutional contours of that question have been the Supreme Court's preoccupation for eighty years. The lodestar is still International Shoe Co. v. Washington, 326 U.S. 310 (1945), which held that due process permits a court to exercise jurisdiction over an out-of-state defendant only where the defendant has "minimum contacts" with the forum such that suit "does not offend traditional notions of fair play and substantial justice." Everything since is gloss on that sentence.

Modern doctrine splits jurisdiction into two species. Specific jurisdiction arises where the lawsuit relates to the defendant's own purposeful contacts with the forum—where the defendant "purposefully availed itself" of the privilege of conducting activities there. The Court has policed both ends of that idea: foreseeability that a product might end up in a state is not enough (World-Wide Volkswagen Corp. v. Woodson, 444 U.S. 286 (1980)), but a single substantial contract deliberately reaching into the forum can be (Burger King Corp. v. Rudzewicz, 471 U.S. 462 (1985)). And in Ford Motor Co. v. Montana Eighth Judicial District Court, 592 U.S. 351 (2021), the Court clarified that the claim need not arise from the defendant's forum contacts in a strict but-for sense; it is enough that the claim relate to those contacts—so Ford, which marketed and sold the same model of car in the forum state, could be sued there over an accident involving one, even though that particular car was first sold elsewhere.

General jurisdiction—the power to hear any claim against a defendant, related to the forum or not—is far narrower, and the modern cases have tightened it dramatically. Under Goodyear Dunlop Tires Operations, S.A. v. Brown, 564 U.S. 915 (2011), and especially Daimler AG v. Bauman, 571 U.S. 117 (2014), a corporation is subject to general jurisdiction only where it is "at home"—paradigmatically its state of incorporation and its principal place of business, and only in the truly exceptional case anywhere else. Daimler buried the older notion that "continuous and systematic" in-state business, standing alone, made a corporation suable anywhere on anything. For an assessor, the practical consequence is concrete: a plaintiff who wants to sue Veridian in a plaintiff-friendly distant forum must show that this dispute connects to that forum, not merely that Veridian does some business there.

In federal court, all of this runs through Federal Rule of Civil Procedure 4(k)(1)(A), which borrows the forum state's long-arm statute—so the analysis is always two steps: does the state's long-arm statute reach this defendant, and does the Constitution permit it? (Some states, like California under Civil Procedure Code § 410.10, extend their long-arm to the full constitutional limit, collapsing the two steps into one.) A separate threshold, subject-matter jurisdiction, asks whether a federal court may hear the case at all—via federal-question jurisdiction under 28 U.S.C. § 1331 or diversity jurisdiction under 28 U.S.C. § 1332, the latter requiring complete diversity of citizenship and an amount in controversy exceeding $75,000. (The federal litigation machinery is mapped in more detail in our comprehensive guide to federal civil litigation for small businesses.)

Even where jurisdiction exists, venue asks whether this district is a proper one. Under 28 U.S.C. § 1391, venue generally lies where any defendant resides (if all reside in the same state) or where a substantial part of the events giving rise to the claim occurred. And a forum-selection clause—the kind of boilerplate parties sign without reading—can override the default analysis entirely. Under Atlantic Marine Construction Co. v. U.S. District Court for the Western District of Texas, 571 U.S. 49 (2013), a valid forum-selection clause is enforced through a transfer motion under 28 U.S.C. § 1404(a) and is given "controlling weight in all but the most exceptional cases," with the plaintiff's ordinary choice of forum getting no weight at all. For Veridian, the single most valuable paragraph in the supply contract may be the forum-selection and dispute-resolution clause, which can dictate not only where the fight happens but whether it happens in court at all—a question that bleeds directly into the choice between litigation and arbitration or mediation.

2. The statute of limitations: has the clock already run?

No defense is cleaner than a winning statute-of-limitations defense, because it disposes of the entire case without a single fact of the underlying dispute being tried. Every civil claim carries a limitations period—commonly two to six years for contract and tort claims, but ranging widely by jurisdiction and claim type—running from the date the claim "accrued." The first task in assessment is to identify the correct period for each claim (a single complaint often pleads several, each with its own clock) and to pin down the accrual date. Accrual is where the genuine analysis lives. The default rule is that a claim accrues when the wrong occurs, but the discovery rule, applied in many jurisdictions and to many claim types, delays accrual until the plaintiff knew or reasonably should have known of the injury and its cause—which can resurrect a claim that looks stale on its face. Layered on top are doctrines that toll (pause) the clock: fraudulent concealment by the defendant, the plaintiff's minority or incapacity, the pendency of related proceedings, and contractual tolling agreements the parties may have signed.

For a defendant, limitations is a defense to assert affirmatively and early—it can support a motion to dismiss where the bar appears on the face of the complaint, or a summary-judgment motion where it requires a few undisputed dates. For a plaintiff, limitations is a constraint to clear before a dollar is spent, and a reason that the pre-suit "golden hour" can be perilously short: a claimant who spends a year exploring resolution may discover that the year cost it the lawsuit. Note, too, that parties can shorten the period by contract—many commercial and employment agreements contain enforceable provisions cutting the limitations window to one year or less—so the contract, not just the statute, must be read.

3. Collectibility: can a judgment ever actually be collected?

This is the question that separates lawyers who think like litigators from lawyers who think like creditors—and it is the question most often skipped, to the client's eventual fury. A judgment is not money. It is a piece of paper that entitles the holder to try to extract money from a defendant who, by definition, did not want to pay. If the defendant is insolvent, judgment-proof, shielded by asset-protection structures, or domiciled where the judgment cannot practically be enforced, then a brilliant liability case and a generous damages award may be worth precisely nothing—minus the hundreds of thousands of dollars spent winning them.

A plaintiff-side assessment therefore must, before committing to suit, investigate the defendant's financial wherewithal: its solvency and cash position, the existence and value of attachable assets, its insurance coverage (often the only realistic source of recovery, which makes coverage analysis part of case assessment, not an afterthought), the priority of other creditors, and the presence of holding-company structures, offshore arrangements, or homestead and exemption protections that can frustrate collection. Where the target is one entity in a corporate family, counsel must consider whether the assets sit in a different entity entirely and whether a veil-piercing or alter-ego theory is realistic—a frequent issue when adversaries run multiple businesses through holding and operating companies. The flip side is just as strategic on the defense: a defendant evaluating settlement should understand its own collectibility profile, because a plaintiff who realizes that a judgment will be uncollectible has a powerful incentive to take a smaller sum in cash today. Collectibility is not a post-judgment problem. It is a pre-filing input, and a case worth a million dollars on the merits is worth zero if the defendant is a shell.

Quantifying the case

With the merits dissected and the threshold gates cleared, the assessment can finally put a number on the case. Traditional valuation leans on qualitative judgment—experienced litigators weighing the strength of claims, the credibility of witnesses, the friendliness of the venue. That judgment is indispensable, but it has well-documented limits: even seasoned attorneys are systematically overconfident, cognitive biases distort their estimates, and "we have a strong case" hands the client a conclusion rather than an analysis it can interrogate. Quantitative methods address these limits not by replacing judgment but by disciplining it—making assumptions explicit, forcing internal consistency, and producing a defensible number.

The foundational method is decision tree analysis, used in commercial litigation for over forty years and accepted by courts, mediators, and funders as a standard valuation methodology. The method has four steps: map the decision points and uncertainties (will the motion to dismiss succeed? will the case survive summary judgment? what are the possible trial outcomes?); estimate the financial consequence at the end of each branch; assign an honest probability to each uncertain event, reflecting the team's real assessment rather than its advocacy; and roll the tree back from the terminal outcomes to the root, multiplying outcomes by probabilities and summing, to produce the case's expected value—the risk-adjusted figure a rational party would accept or pay to resolve it.

Veridian's case illustrates the mechanics. Suppose its team assesses a 40% chance that its motion to dismiss succeeds (outcome: $0 in damages, $50,000 in fees). If the motion fails, a 50% chance of winning summary judgment (outcome: $0 damages, $200,000 in cumulative fees). If that fails, a 45% chance of a defense verdict at trial ($0 damages, $500,000 in cumulative fees) and a 55% chance of losing ($4 million in damages plus $500,000 in fees). Rolling back: the expected cost of trial is (0.45 × $500,000) + (0.55 × $4,500,000) = $2,700,000; factoring in summary judgment, the expected cost if the motion to dismiss fails is (0.50 × $200,000) + (0.50 × $2,700,000) = $1,450,000; and factoring in the motion to dismiss, the overall expected cost is (0.40 × $50,000) + (0.60 × $1,450,000) = $890,000. On a risk-adjusted basis, the case is worth roughly $890,000—so a settlement below that is economically favorable and one above it is not.

The figure is only as good as its inputs, which is why sensitivity analysis (varying the key probabilities to see how the value moves) and stress testing (asking what the case is worth if every close call goes against the client) matter. If the value swings dramatically on one assumption, that assumption becomes a priority for investigation; if it is stable across a range, the settlement analysis is robust. Notice, too, that this is exactly where the liability-versus-damages split pays off: the decision tree forces counsel to separate the probability of losing (a liability question) from the magnitude of the loss (a damages question), so that the team is never lazily multiplying a single fuzzy number against the demand.

The expected value does more than name a figure; it defines a settlement zone and a negotiating posture. Veridian's $890,000 expected cost is the point of indifference for a risk-neutral defendant, but most clients are risk-averse, and the value of eliminating uncertainty may justify paying somewhat more than the pure expected value to settle—just as the distributor, if it is risk-averse or cash-constrained, may accept somewhat less than its own expected recovery. The gap between what each side would rationally accept is the settlement zone, and a well-built decision tree lets Veridian estimate not just its own number but, by modeling the case from the plaintiff's side, where the plaintiff's number likely sits. That dual modeling is what turns a static valuation into a negotiating tool: it tells Veridian whether a given demand falls inside or outside the rational settlement zone, what counteroffer is defensible, and which uncertainties—if resolved through a targeted motion or a key deposition—would move the plaintiff's number most. The tree is rebuilt as the case develops and the probabilities sharpen; the settlement zone narrows as discovery resolves the open questions, and the moment a decisive ruling lands—a surviving motion to dismiss, a favorable claim construction, a key admission—the recalculated expected value tells the client immediately how its posture should change.

Risk tolerance: why two identical cases settle differently

The expected value is a number, but the decision it informs depends on something the number cannot capture: the client's tolerance for risk. Two clients facing identical claims may rationally choose opposite strategies—one settling early to eliminate uncertainty, the other litigating aggressively to establish a precedent—because risk tolerance, financial position, and strategic objectives differ. A $4 million verdict is an existential threat to a startup and a rounding error to a Fortune 100 company; a business in a heavily regulated industry calculates risk differently from one in an unregulated market; a company repeatedly targeted by similar claims may develop a higher tolerance and a strategic preference for fighting; and a consumer-facing brand may weight reputational exposure above pure economics. The assessment must translate the expected value into a recommendation calibrated to this client.

Two refinements matter. First, sophisticated clients managing large litigation portfolios often think about risk not case by case but across the portfolio: a company that would settle any individual $2 million claim may nonetheless litigate aggressively across its docket to deter future claims, accepting occasional trial losses because the deterrent posture reduces the total number of claims over time. For such a client, Veridian's distributor suit is assessed not only on its own expected value but on the signal its resolution sends to other distributors watching to see whether a similar claim would pay. Second, the analysis should be framed as return on investment—not merely "what will this cost," but "will the company avoid more costly problems by investing in this litigation, or will its position worsen if it does not terminate the exposure early?" Failing to assess a case properly is itself a risk, as real as overspending, and the discipline is to identify, monetize, and weigh that risk rather than defaulting to whichever instinct—fight or settle—feels safer in the moment.

Where predictive analytics fit—and where they don't

The traditional weakness of decision trees is that the probability inputs are subjective: two equally experienced lawyers may assign meaningfully different odds to the same event, with no objective basis for choosing between them. AI-powered analytics platforms are starting to change that. Tools that analyze large volumes of court records—judicial rulings, motion outcomes, case durations, damages awards—can forecast the likelihood that a particular judge grants a motion to dismiss, estimate a case's expected duration in a given venue, identify historical win rates for a claim type, and suggest when similar cases tend to settle. The major platforms report meaningful accuracy on discrete questions like dismissal prediction, drawing on tens of millions of court documents.

These tools are most valuable at the early-assessment stage, where they can inform venue analysis, motion strategy, settlement timing, and budget forecasting. But two cautions are essential. First, they do not replace judgment: a model cannot weigh case-specific facts, witness credibility, or the unpredictable dynamics of a particular dispute, and the most sophisticated model is only as good as the qualitative assessments that feed it. Second, the field is evolving quickly and unevenly, and reported accuracy figures describe narrow, well-defined questions rather than holistic case outcomes. The right posture is to use predictive analytics as an objective check on subjective probability estimates—a way to test whether the team's gut number for "this judge grants this motion" aligns with the historical record—not as an oracle. Used that way, the analytics sharpen the decision tree rather than substituting for it.

Who pays, and on what terms: fee arrangements and the ethics of the engagement

An assessment that quantifies exposure but ignores the cost of the lawyering is half an assessment. The fee arrangement is not a back-office detail; it shapes incentives, allocates risk between client and counsel, and is itself governed by rules of professional conduct that a careful lawyer treats as part of the engagement, not paperwork to be skimmed.

The traditional billable hour—fee equals hours worked times rate—remains the default for large, complex, or novel matters, where the work is genuinely unpredictable. Its defect is notorious: it ties compensation to effort rather than outcome, and so it neither rewards efficiency nor aligns counsel's interest with the client's. The client bears all the risk of delay, overrun, and a bad result. In response, alternative fee arrangements (AFAs) have proliferated, and they are limited, as the saying goes, only by the creativity of the lawyer and client who design them—subject always to the ethical ceiling discussed below. The principal structures:

  • Flat or fixed fees set a single price for a defined scope—a task ("draft and file the motion to dismiss"), a litigation phase ("all of fact discovery"), an entire matter, or a portfolio ("all of the company's warranty litigation in this jurisdiction for the year"). The firm bears the overrun risk; the client gains cost certainty and a lighter administrative load. The danger is the inverse: a firm that finishes early reaps a windfall, and one that underpriced may be tempted to cut corners. Flat fees suit commoditized, repeatable work; they suit one-off, bet-the-company matters poorly.
  • Contingency fees pay counsel only on recovery, as a percentage of the amount recovered—classically about one-third in personal-injury work, though there is no fixed "reasonable" percentage. They open the courthouse to plaintiffs who could not otherwise afford it, and they shift risk to counsel. The defense-side mirror image is the reverse contingency, where the fee turns on how far below an agreed "fair value" of the claim counsel holds the plaintiff's recovery.
  • Capped and collared fees are hourly billing with guardrails: a "cap" sets a ceiling on total fees, a "collar" pairs that ceiling with a floor, and a "soft cap" pays counsel a reduced rate for work beyond the cap rather than nothing.
  • Holdbacks and success fees layer onto any of the above—the client withholds a portion of the fee (often around 20%) until defined milestones are met, and pays a premium for an exceptional result (early resolution, a result well under budget, a win that exceeds expectations).
  • Hybrids—a discounted hourly rate plus a contingency slice, for instance—split the difference, giving counsel some guaranteed compensation while preserving an outcome-based upside and sharing risk both ways.

Every one of these structures runs into ABA Model Rule 1.5, which governs fees in nearly every U.S. jurisdiction (in some adapted form). Rule 1.5(a) forbids an "unreasonable" fee and lists the factors that bear on reasonableness—the time and labor required, the novelty and difficulty of the questions, the customary fee in the locality, the amount involved and results obtained, time limitations, the nature of the professional relationship, and counsel's experience and ability, among others. Rule 1.5(b) requires that the basis of the fee be communicated to the client, preferably in writing, before or within a reasonable time after the engagement begins. Rule 1.5(c) requires that contingent-fee agreements be in writing, signed by the client, stating the method by which the fee is determined; and Rule 1.5(d) flatly prohibits contingent fees in two settings—domestic-relations matters where the fee is contingent on securing a divorce or on the amount of alimony, support, or property settlement, and the representation of a criminal defendant. The practical upshot for assessment is twofold: a fee structure must be chosen with the client's risk profile and cash position in mind (cost certainty argues for a flat fee, an uncertain but quantifiable recovery for a contingency or hybrid), and the chosen structure must be papered in a retainer that satisfies Rule 1.5—because a fee the client did not knowingly agree to, or one a court later deems unreasonable, is a fee the lawyer may not be able to collect.

The other ethical rule that belongs in every intake assessment is Model Rule 1.7, on concurrent conflicts of interest. A conflict exists if the representation of one client is directly adverse to another client, or if there is a significant risk that the representation will be materially limited by counsel's responsibilities to another client, a former client, or a third person—or by counsel's own personal interest. Some conflicts are consentable: the client may waive them if the lawyer reasonably believes competent, diligent representation can still be provided to each affected client, the representation is not prohibited by law, the clients are not asserting claims against each other in the same proceeding, and each affected client gives informed consent, confirmed in writing. Others are non-consentable—most obviously, a lawyer cannot represent both sides of the same lawsuit no matter how enthusiastically everyone consents. The point for assessment is that the conflicts check is not a clerical box to tick after the engagement letter is signed; it is a substantive gate that can disqualify counsel, delay the case, and—if missed—blow up the representation midstream. (An at-risk fee creates its own subtle Rule 1.7 tension worth flagging to the client: counsel with a financial stake in a quick recovery may feel a pull toward early settlement that does not perfectly track the client's objectives.) For a fuller map of who does what and how lawyers are engaged, see our overview of the types of lawyers.

E-discovery: proportionality and technology-assisted review

Electronically stored information has transformed litigation budgets, and for many cases the cost of identifying, preserving, collecting, processing, reviewing, and producing data exceeds every other litigation cost combined. The 2015 amendments to Federal Rule of Civil Procedure 26(b)(1) made proportionality a central limit on discovery: it must be "proportional to the needs of the case," weighing the stakes, the amount in controversy, the parties' relative access to information, their resources, the importance of the discovery, and whether its burden outweighs its likely benefit. Proportionality is not only a shield against the opposing party's overreach; it is an affirmative discipline that should shape the client's own practices. Early in the case, counsel should map the likely scope of e-discovery—the custodians, the systems, the data volume, the estimated review cost—because that analysis drives the budget, the preservation strategy, and the negotiations with opposing counsel and the court.

The first concrete step is identification: determining what data exists, where it lives, who controls it, and what preservation is required. This means mapping the client's information architecture—email systems, document-management platforms, collaboration tools like Slack, Teams, and SharePoint, cloud storage, mobile devices, and legacy systems—because relevant data hides in places a custodian interview alone will not surface. For Veridian, the supply-contract dispute may turn on messages exchanged in a procurement chat channel or revisions tracked in a shared document, neither of which lives in the email system the company would instinctively search first. A litigation-hold and collection plan that misses those repositories does not just risk spoliation sanctions under Rule 37(e); it risks building the entire assessment on an incomplete factual record.

The single biggest cost driver has historically been document review—often 70% or more of the e-discovery budget. Technology-assisted review (TAR), or predictive coding, has dramatically reduced it by training a machine-learning algorithm on a sample of attorney-reviewed documents and applying the learned patterns to the full population. Studies consistently show TAR is as accurate as—and often more consistent than—exhaustive human review, at a fraction of the cost and time, and courts have broadly accepted it as defensible, sometimes ordering its use where manual review would be disproportionately expensive. The practical consequence for assessment is that cases involving large document populations, once prohibitively expensive, may now be economically viable—and that reality belongs in the settlement calculus. Throughout, privilege protection must be built into the workflow from the start—privilege search terms, senior-attorney review of flagged documents, and quality-control sampling—because in a review of hundreds of thousands of documents even a small error rate can produce inadvertent production of privileged material; while Federal Rule of Evidence 502(b) protects against waiver for inadvertent disclosures where reasonable steps were taken, the safest course is to prevent the disclosure in the first place.

Researching the adversary

Effective assessment requires intelligence about the whole competitive landscape. Research the assigned judge—how often they grant motions to dismiss, how they handle discovery disputes, whether they push early settlement, how long they take to rule—work that AI judicial-analytics platforms have made dramatically easier. Research opposing counsel—whether they favor aggressive or proportional discovery, whether they typically try cases or settle, whether they have the resources to fund extended litigation, and, critically, whether they are backed by a litigation funder. And research the opposing party—its financial condition (which, as we saw, doubles as collectibility analysis), litigation history, public statements, and, for corporate adversaries, SEC filings and earnings calls for statements bearing on the dispute. For Veridian, learning that the distributor's counsel is funded and has the patience to try the case would materially change the settlement calculus.

Witnesses and documents

Identifying key witnesses is not merely listing the people who know something; it requires assessing each along several dimensions—what they know, how credibly they can communicate it, whose interests they align with, and how the opposing party will use them. For each potential witness, counsel should ask whether they are friendly, neutral, or hostile; whether they will be available at trial given employment status and location; whether their testimony is corroborated by documents (which strengthens it) or contradicted by them (which makes it a liability); and whether they carry impeachment material—prior inconsistent statements, criminal history, bias—that could undermine them. Expert witnesses require separate assessment: the subject areas where expert testimony will be needed, the qualifications required, the likely cost, and the availability of experts who can survive a Daubert challenge to their methodology. For Veridian, the former sales manager who negotiated the supply contract may be the most important witness in the case—and whether he is now a loyal employee, a neutral retiree, or a disgruntled former employee who left for a competitor may matter more to the case's value than any legal argument.

Document identification is the mirror image. In the digital environment, "documents" span emails, text messages, chat platforms, social posts, financial records, meeting minutes, contracts, system logs, metadata, and database records, and the assessment must surface both the documents that help the client and the ones that hurt. It is always better to find the damaging document early—when a strategy for addressing it can be developed—than during trial preparation, when options have narrowed. The proof matrix and the document inventory work together: the matrix says what must be proven, and the inventory says what evidence exists to prove or disprove it, with the gaps between them defining the discovery plan.

Third-party litigation funding and the assessment

Third-party litigation funding (TPLF) has become a significant factor in case assessment on both sides of the docket, and its disclosure rules are in flux—so practitioners need the current state, not last year's. The U.S. commercial litigation-finance industry manages billions in assets, and funded plaintiffs increasingly appear across commercial and patent litigation; in patent cases in particular, a substantial share of suits against operating companies involve confirmed or suspected outside funding.

The presence of funding changes the assessment in concrete ways. For a defendant like Veridian, a funded opponent may have the resources to pursue protracted litigation, retain top experts, and resist the settlement pressure that ordinarily works against an under-resourced plaintiff—and the funder's economic interest may shape the plaintiff's settlement authority, since some funding agreements require the plaintiff to consult the funder before accepting an offer. For a plaintiff, funding can make otherwise unaffordable litigation viable, but at a price: the funder typically takes a meaningful percentage of any recovery, reducing the plaintiff's net, so the assessment must account for the funder's return expectations.

The disclosure rules are where the law is moving fastest. At the federal-rules level, the Advisory Committee on Civil Rules has been studying whether to require disclosure of funding arrangements; at its October 2025 meeting it agreed to continue studying the question rather than propose a rule, and in early 2026 advocates submitted proposed language to amend Rule 26(a)(1)(A) to require disclosure of a nonparty funder's identity and the funding agreement at the outset of a case—so a federal rule remains under active consideration but is not yet adopted. In Congress, several bills are pending, including measures that would require disclosure of TPLF in federal civil cases and impose disclosure obligations and limits on funder control in class actions, MDLs, and large coordinated proceedings, with a separate strand targeting foreign funding specifically. At the state level, the patchwork has expanded well beyond the early adopters—a growing number of states have enacted or are considering mandatory disclosure regimes, several with provisions aimed at foreign funders—and the insurance industry has introduced an optional commercial-liability endorsement allowing either party to demand mutual disclosure of funding as a condition of coverage. The net picture as of mid-2026 is that mandatory federal disclosure has not yet arrived but is closer than ever, while a thickening state and contractual patchwork already requires disclosure in many forums. Counsel must monitor these developments, because disclosure obligations materially affect the strategic calculus for funded plaintiffs and the defendants who face them.

From game plan to budget

After the preliminary assessment, counsel builds a strategic framework mapping the best options at each stage. This starts with issue-spotting—the type of case, exposure to punitive damages or statutory penalties, applicable damages caps, the risk of copycat litigation, insurance-coverage questions, and whether the dispute attacks the client's business model. It proceeds to the strategic decision points: can the case settle, or does settling create unacceptable precedent? Could litigating or settling invite further suits? Is there related litigation to coordinate? Could a structured conversation between executives resolve the dispute outside litigation? And it weighs early ADR—mediation (most effective once both sides have a realistic view of their positions), early neutral evaluation (a respected neutral's non-binding read that can anchor settlement), and arbitration—against continued litigation, while accounting for the non-litigation costs that never appear on a budget but are often the largest: reputational harm, damaged business relationships, executive distraction, and collateral regulatory or criminal exposure. A case assessment that ignores these indirect costs will systematically undervalue settlement. (The forum question—court, arbitration, or mediation—deserves its own deliberate analysis; see our practical guide to choosing a dispute-resolution forum.)

All of this culminates in a phased litigation budget, organized by stage—pre-filing, pleadings, fact discovery, expert discovery, class certification (if applicable), summary judgment, pre-trial, trial, post-trial, and appeal—with separate estimates for attorney fees, expert fees, e-discovery, depositions, ADR, and other costs. The budget should be a decision tool, not just a cost estimate: at each phase boundary it should include a decision point at which counsel and client reassess whether to proceed, settle, or change course based on what has been learned. That structure keeps the litigation from becoming an autopilot exercise in which costs accumulate without deliberate evaluation, and it gives the client the cost-curve visibility that rational settlement decisions require. The budget and the fee arrangement should be designed together: a phased budget pairs naturally with phased flat fees or capped fees, each phase boundary doubling as both a financial checkpoint and a strategic one.

Testing the case and charting the stages

Because lawyers develop confirmation bias living with a case, mock trials and jury research provide a corrective by introducing the perspective of people with no stake in the outcome. Conducted early—within the first 90 days—they reveal how ordinary people react to the case's themes, identify which witnesses and evidence persuade and which damage, test damages theories, and highlight the factual questions jurors actually care about, which often differ from the legal questions lawyers instinctively prioritize. Where budget is tight, the exercise scales down: focused best-case/worst-case mock sessions, smaller focus groups, online theme-testing with larger samples, or a bench-trial simulation with a retired judge.

Finally, the assessment should chart costs and outcomes stage by stage, comparing at each point the cost and likelihood of favorable termination against the cost and risk of continuing. At the pleadings stage, is a motion to dismiss viable, and what is the realistic probability given the assigned judge's record? In a putative class action, what is the likelihood of defeating certification, and how would the decision reshape settlement dynamics? At summary judgment, how does the judge's grant rate factor in, and how might even a denied motion create settlement leverage by forcing the opponent to confront its weaknesses? At trial, what are the total costs, the realistic probability of a favorable verdict, the damages range, and the reputational exposure? This stage-by-stage view lets the client make informed decisions throughout the case rather than deferring all strategic thinking to the eve of trial.

Special considerations for intellectual property cases

IP disputes warrant particular attention because of their technical complexity, specialized frameworks, and outsized business impact. In patent litigation, claim construction (the Markman hearing, under the standard of Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005)) is often the pivotal event, and counsel should assess the likely construction before investing heavily in infringement or invalidity analysis; the availability of inter partes review at the PTAB—faster and cheaper than district-court litigation, with a lower invalidity burden, but subject to a one-year filing deadline and estoppel consequences—reshapes strategy, while damages analysis must weigh the reasonable-royalty framework of Georgia-Pacific against lost profits and any apportionment requirement. Recall, too, that the heightened clear-and-convincing standard for proving invalidity (Microsoft v. i4i, above) is itself a load-bearing input in any patent-case decision tree. In trademark disputes, the multi-factor likelihood-of-confusion analysis drives the case, with particular attention to the strength of the mark and the availability of actual-confusion evidence (where a well-designed consumer survey can be powerful but costs $100,000–$300,000 or more), and injunctive relief is evaluated under the four-factor eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388 (2006), test. In copyright litigation, assessment must address both copyrightability and the specific rights infringed, the fair-use defense under the four-factor test of Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569 (1994), as refined by Andy Warhol Foundation for the Visual Arts, Inc. v. Goldsmith, 598 U.S. 508 (2023) (which sharpened the focus on whether a use has a genuinely different purpose or character), and the statutory-damages calculus, which—for works timely registered—can reach $750 to $30,000 per work, or up to $150,000 for willful infringement, without proof of actual damages, making timely registration a threshold assessment question. And in trade secret disputes, which often arise from employee departures and carry the urgency of ongoing, irreversible harm, counsel must assess whether the information qualifies as a trade secret, whether the owner took reasonable measures (the most-litigated element), whether there is evidence of actual or threatened misappropriation rather than mere suspicion, and whether emergency relief—a TRO or preliminary injunction, or in extraordinary cases the DTSA's ex parte seizure (18 U.S.C. § 1836)—is available and advisable given its procedural complexity and the risk of a wrongful-seizure claim.

A short FAQ

What is "early case assessment," and how is it different from just answering the complaint? Answering the complaint is a procedural obligation; early case assessment is a strategic exercise. Assessment asks the questions the answer cannot—can each element be proven, under what burden, in what forum, against a defendant who can pay—and produces a defensible valuation and a budget. The answer reacts to the lawsuit; the assessment positions the client to shape it.

How much should case assessment cost? Less than almost anything else in the litigation, and it pays for itself many times over. A focused assessment in a mid-sized commercial dispute might run a small fraction of one phase of discovery, yet it routinely changes whether the client spends seven figures fighting a case it should have settled, or settles a case it should have fought. The expensive mistake is skipping it.

Who decides whether to settle—the lawyer or the client? The client, always. Counsel's job is to present an honest assessment and a set of options with their expected values and risks; the decision to settle, mediate, arbitrate, or try the case belongs to the client. A lawyer who substitutes a conclusion ("we should settle") for an analysis has done the client a disservice—and, where a contingent or other at-risk fee gives counsel a financial stake in the timing, has flirted with a Rule 1.7 problem.

We have a strong case on liability. Why are you still talking about money? Because liability and damages are different questions, and collectibility is a third. A 90%-liability case worth $200,000 against an insolvent defendant is worth far less than a 40%-liability case worth $4 million against a well-insured one. The number that matters is the one that survives all three filters.

Can the statute of limitations really end a case before we get to the facts? Yes—cleanly and completely. A claim filed after its limitations period has run is barred regardless of how meritorious it is, which is why identifying the correct period and accrual date for each claim is among the first things a competent assessment does, and why a plaintiff who spends too long exploring resolution can lose the very right it was trying to vindicate.

Conclusion: assessment shapes outcomes, it doesn't just predict them

Case assessment is the discipline that turns litigation from a reactive, crisis-driven activity into a deliberate, strategy-driven one. It is not merely about predicting outcomes—though that matters—but about shaping them: identifying the client's true objectives, dissecting each claim element by element under the governing burden of proof, separating liability from damages, clearing the threshold gates of jurisdiction, limitations, and collectibility, quantifying the range of outcomes, and supplying the framework through which every later decision gets made. The modern litigator's tools—decision trees, expected value, AI-powered analytics, technology-assisted review, judicial analytics—have made assessment more rigorous and more defensible than ever, but technology does not replace judgment; the most sophisticated model is only as good as the qualitative assessments that feed it, and the most detailed budget is only as useful as the strategic framework that gives it context.

The most effective litigators combine analytical rigor with strategic creativity—using quantitative tools to test their qualitative judgments, challenging their own assumptions through red-team exercises and mock trials, and presenting clients with honest assessments and clear options rather than conclusory reassurance. For Veridian, the difference between forwarding the complaint and waiting, on the one hand, and conducting a disciplined early assessment, on the other, is the difference between a case that manages the company and a company that manages the case. In an environment of rising costs, growing third-party funding, and diminishing judicial patience for inefficiency, the litigator who invests in thorough early assessment is not spending money. They are saving it.

For assistance assessing litigation risk and building a case strategy, contact our litigation practice.

Related Articles

Selected Authorities

International Shoe Co. v. Washington, 326 U.S. 310 (1945); World-Wide Volkswagen Corp. v. Woodson, 444 U.S. 286 (1980); Burger King Corp. v. Rudzewicz, 471 U.S. 462 (1985); Goodyear Dunlop Tires Operations, S.A. v. Brown, 564 U.S. 915 (2011); Daimler AG v. Bauman, 571 U.S. 117 (2014); Ford Motor Co. v. Montana Eighth Judicial District Court, 592 U.S. 351 (2021); Atlantic Marine Construction Co. v. U.S. District Court for the Western District of Texas, 571 U.S. 49 (2013); Microsoft Corp. v. i4i Ltd. Partnership, 564 U.S. 91 (2011); Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005); eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388 (2006); Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569 (1994); Andy Warhol Foundation for the Visual Arts, Inc. v. Goldsmith, 598 U.S. 508 (2023). 28 U.S.C. §§ 1331, 1332, 1391, 1404(a); Fed. R. Civ. P. 4(k), 26(b)(1), 37(e); Fed. R. Evid. 502(b); 18 U.S.C. § 1836 (DTSA); ABA Model Rules of Professional Conduct 1.5 and 1.7. TPLF disclosure measures remain under active consideration at the federal-rules level and in Congress as of mid-2026; several are pending rather than enacted.

This article is for general informational purposes only and does not constitute legal advice, nor does it create an attorney-client relationship. Procedural rules, limitations periods, the law of personal jurisdiction and litigation funding, the rules of professional conduct as adopted in each jurisdiction, and the available technologies all vary by jurisdiction and continue to evolve—several of the funding-disclosure measures described are pending rather than enacted. Consult the applicable rules and qualified counsel about your specific matter.