Posted by Rachel Coleman, senior legal industry specialist at Legal Futures Associate iManage [1]
Coleman: Value-based pricing appeals to clients
For years, the legal profession has talked of moving away from the billable hour to value-based pricing without quite managing it.
Value-based pricing is viewed as the commercially sensible way to avoid a ‘race to the bottom’ as AI drives efficiency across legal, delivery costs fall and corporate legal teams outsource less legal work.
However, as AI platforms shift toward consumption-based pricing, the cost of delivering that same work will increase.
Does this shift threaten the value-based pricing movement?
Rethinking how AI costs fit into legal pricing
AI platforms have typically been priced on a flat, per-seat basis, which means that the cost of these platforms has been predictable, if substantial.
At present, the cost of AI often sits inside the IT or wider innovation budget, folded into the cost of doing business. That cost may feed through into charge-out rates in the ordinary way that firm overheads do but masks the urgency to identify return on the AI investment.
That changes as vendors move toward consumption-based pricing, charging by usage rather than by user. What was an accepted fixed cost becomes a variable challenge.
Firms could cap usage of their AI systems, with individual or team allowances and an escalation process for exceeding them.
But legal work is always uneven: a large due diligence exercise can burn through an individual allowance in hours, raising immediate questions about project-level budgets and whether firms need a return-on-investment case before releasing additional AI spend on a matter that still needs to be addressed.
Further, passing AI costs directly to clients as a disbursement is not gaining traction quickly in the UK.
Clients have accepted SaaS-style charges for platforms they use directly, but a line item for ‘AI used on your matter’ is a different proposition entirely, not least because that cost could be uncapped at engagement until such a point in the future that firms have sufficient data to accurately predict AI costs for certain matters.
A workable route remains true value-based pricing: pricing on the output and outcome delivered to the client, rather than the hours or tools used to get there, so that any additional AI cost is absorbed within the price the client has already agreed.
Value-based pricing offers a stronger shield against rising AI costs – but only if it is applied with discipline.
Fixed-based pricing arrived at from an underlying cost of doing business will be vulnerable to gross margin erosion (whether AI costs are allocated to the matter directly or not) until that time firms can accurately predict the cost of AI.
For clients, the appeal is straightforward: whilst value-based pricing may not mean cheaper, it keeps the invoice predictable and the legal services provided hyper focused on delivering client value (speed, accuracy, insight, market knowledge) without disincentivising AI adoption at a time when clients are demanding firm innovation.
Structure guards margins against erosion
Even when legal services are value priced, indiscriminate AI usage could erode gross margin.
So, lawyers must firstly be educated on the cost implications of the tools they use – reinforcing the message that AI is not ‘free’ and it carries direct cost consequences that can escalate easily.
Two practices help here: triaging work before defaulting to AI, just as you would when allocating it to a team, so AI is used where it adds value rather than by default; and matching model choice to task complexity, using a lighter (and cheaper) model wherever it does the job just as well.
Secondly, for activities that run repeatedly across the team, lawyers should build repeatable workflows where possible, rather than starting from scratch on every matter, driving consistency and steadying token spending.
Thirdly, firms and vendors should work together to ensure that vendors can surface sufficient data on where their AI is being used, by whom and for what, to enable central firm teams to monitor AI usage and map the costs associated with that usage against matters.
Ready for what’s next
It is not hard to imagine a distinction emerging between business-as-usual AI costs and high-value AI use.
Routine AI activity that supports multiple matters may simply become part of the firm’s accepted cost base, with repeatable workflows making it as efficient as possible and costs as predictable as possible.
By contrast, large-scale or matter-specific AI use may justify allocating AI costs more directly to the matter. Even then, it feels unlikely that firms will pass those costs through as an uncapped disbursement to their clients; robust value-based pricing is a better route to a fair outcome for both firm and client.
However the work evolves, firms will need a more deliberate approach to pricing if they want their economics to keep pace with the way matters are delivered in the AI era.