Introduction
Harvey has traditionally sold its platform on a per-seat basis: a firm or legal department pays for a number of lawyers who can use Assistant, Vault, Knowledge and Workflows. As Harvey Agents begin to execute legal work end to end (for example reviewing an entire data room, drafting a first-pass contract set, or preparing a discovery summary), the value delivered per seat varies enormously. One associate might run agents that do the work of a team for a week; another might ask a few questions a month.
The prompt: "Should Harvey price its agents per seat, as it does today, or move to usage- or outcome-based pricing?" This trade-off affects revenue, customer adoption, margins and how law firms bill their own clients.
I'll ask clarifying questions, classify the trade-off, analyse the product, form a hypothesis, choose metrics, design an experiment, plan the analysis, set a decision framework and recommend next steps.
Step 1
Clarifying Questions (3 minutes)
Why it matters: If costs vary widely, flat seat pricing can create unprofitable heavy users. Expected answer: A large diligence agent run can cost 100x a typical chat interaction. Impact on approach: Pure seat pricing risks margins for heavy agent workloads.
Why it matters: Law firms and legal departments prefer predictable annual costs. Expected answer: Annual budgets; surprise bills are strongly disliked. Some firms want to pass AI costs through to clients on specific matters. Impact on approach: Any usage component must be predictable and matter-attributable.
Why it matters: Seats may keep agents away from paralegals and staff who could use them. Expected answer: Yes, firms often license only a subset of lawyers. Impact on approach: Usage-based options could expand reach.
Why it matters: Matter-level billing requires reliable attribution. Expected answer: Usage can be tagged to Vault projects and matters. Impact on approach: Matter-level pricing is feasible.
Practice similar questions
Subscribe to access the full answer
.png){kind=logo}
