Pricing the outcome rather than the hours solves the AI problem and introduces a harder one: you now have to be right about scope.
Start a conversation with the AI Adoption Concierge, already scoped to alternative fee arrangements. Pick a starting point, or describe your situation directly.
Alternative fee arrangements are the obvious response to work that has become dramatically faster, and firms are expanding flat-fee and value-based pricing for precisely that work. The appeal is straightforward — decouple revenue from hours and the compression stops being a threat. The difficulty is equally straightforward and much less discussed: a flat fee requires the firm to predict scope accurately, absorb the variance when it is wrong, and have cost data good enough to price from. Firms whose only instrument has been the hour are frequently missing all three, and the first year of alternative pricing is where they find out.
Each shifts risk somewhere. The question is whether the firm can carry the risk it has just taken on.
Simplest for the client. Requires accurate scope prediction and the discipline to enforce scope.
Hourly with a ceiling. Keeps the familiar mechanics while giving the client certainty; the firm absorbs the tail.
Ongoing access for a recurring fee. Suits advisory relationships; needs a defined boundary or it becomes unlimited.
Priced by phase with a decision point between. Reduces the scoping risk by shortening the horizon.
Priced to what it is worth. Powerful and demanding, and constrained by rules on contingency in many contexts.
A defined deliverable at a published price — the structure AI makes newly viable for routine work.
What has to be in place first.
It moves risk from the client to the firm. That is the point, and also the exposure.
A firm that prices by value while measuring and rewarding its lawyers by hours has set the two in opposition. The people billing are being asked to become more efficient at the thing their compensation counts. Pricing changes that skip this stall quietly.
With high-volume, well-understood matter types where the firm has real historical cost data and the scope is genuinely predictable — the work AI compresses most is usually also the work that prices most safely. Starting with complex, bespoke matters is the common instinct, because that is where the fees are largest, and it is where flat pricing goes wrong fastest. Beginning narrow also gives the firm real data on variance before committing more of the book.
This is exactly the point on which jurisdictions have diverged, and it is a question for the firm's own counsel rather than a general rule. Virginia has indicated that keeping a non-hourly fee unchanged for AI-assisted work is not per se unreasonable, on the reasoning that a flat fee prices the outcome. ABA Formal Opinion 512 is more cautious, warning a flat fee may be unreasonable where AI drastically reduced the effort. A firm across several states should not assume one jurisdiction's view travels.
Write the scope down in terms a client would recognise, define what falls outside it, and agree the mechanism for handling change before the matter starts rather than during it. The failure is rarely a single large excursion; it is accumulated small accommodations nobody wanted to charge for. Firms that price flat successfully generally have someone other than the relationship partner watching scope, because the person who owns the relationship is the worst-placed to raise it.
They stop measuring what the firm now wants, which is a governance problem more than a pricing one. If revenue on a matter is fixed, hours spent are a cost rather than a product, and a lawyer optimising for recorded hours is optimising against the firm. Firms moving meaningfully to alternative pricing generally have to change how contribution is measured — toward matter profitability, client outcomes and origination. That is a harder internal conversation than the pricing itself, and it is where most attempts stall.
Describe the matter types you would start with. The Institute will help you assess whether you can price them.