In this piece for Artificial Lawyer, Darrow's COO Mathew Keshav Lewis makes the case that plaintiff firms need to manage their dockets the way sophisticated investors manage portfolios — and that AI is finally making that level of discipline achievable.
Lewis opens with a simple but powerful analogy: the best financial investors don't win every bet. They win by managing risk, timing, and liquidity across a portfolio. That same logic applies to contingency-fee plaintiff firms, which have historically operated by betting heavily on a small number of high-profile cases — only to face serious financial strain when those cases stall, settle poorly, or collapse entirely. The result is a profession that is far more exposed to volatility than it needs to be, given how much data now exists about case outcomes, durations, and settlement patterns.
The article introduces four dimensions of portfolio diversification that firms can apply to their caseloads. The first is timing: cases resolve on different timelines, and a docket composed entirely of long-duration matters creates severe liquidity pressure, forcing firms to carry costs for years before any payout arrives. Pairing quick-turn cases with longer-running ones creates a more predictable cash flow waterfall. The second dimension is settlement value: a mix of high-, mid-, and low-value cases balances stability against growth potential. Too many small cases limits revenue growth; too many moonshots almost guarantees cash droughts between payouts.
The third dimension is practice area diversification. Just as investors avoid overconcentrating in a single sector, firms that operate primarily in one area of law are exposed to adverse rulings or regulatory shifts that can damage an entire practice overnight. Building referral networks and business development partnerships across case types reduces that risk meaningfully. The fourth dimension is conviction weighting — moving beyond intuition when deciding whether to take a case. Using probability-adjusted outcome modeling based on fact strength, jurisdiction, defendant solvency, and historical settlement data allows firms to allocate resources in proportion to the likely return.
Lewis then addresses how AI changes the economics of applying these principles. The underlying data has always existed in theory — case outcomes, settlement sizes, duration curves, win rates — but it was too fragmented and voluminous for any firm to synthesize meaningfully on its own. AI tools now make it possible to build real, forward-looking cash flow models from this data, helping firm partners see their financial health across the full docket and identify periods where reserves may fall short of operational costs before those periods arrive.
He envisions what a modern plaintiff firm management dashboard could look like: portfolio composition by practice area and settlement value, quarterly cash flow forecasts, settlement trends by jurisdiction, optimal claimant counts for mass arbitrations, and simulated portfolio outcomes under different case mix scenarios. This is precisely what Darrow provides — the analytics and legal intelligence infrastructure to help firms run their dockets with the discipline of an investment portfolio.
The piece closes with a strategic argument: firms that adopt portfolio theory will not only run more stable businesses but will attract capital partners who understand data-driven management. The best plaintiff firms, Lewis argues, will use these tools to compound both financial returns and justice.
