A faster first draft is a useful operational result. It is not, by itself, a reason to cut a fee or evidence that a matter became more profitable. For a law firm owner, the next question is what changed across the whole service: preparation, review, revisions, client communication, and final delivery.
That distinction matters as AI moves into pricing conversations. The practical response is to build evidence around one service before changing the firm's broader fee structure.
What the latest research says about pricing pressure
In its 2026 legal report, Thomson Reuters reports that 71% of in-house legal professionals expect outside firms to change their commercial models as AI use increases, while 28% of law firms report pricing changes in response to AI. The research used March and April 2026 responses from 736 law firm professionals across 46 countries, with additional references to 203 corporate legal respondents globally. These are survey findings about expectations and reported actions, not a forecast for every local practice. Read the report and methodology.
Clio's 2025 Legal Trends Report adds useful historical context: 59% of firms in its analysis billed flat fees, either exclusively or alongside hourly fees, in 2024. That does not establish that AI caused the shift or that a flat fee suits every matter. See Clio's pricing analysis.
Start with one service and a written scope
For a small firm, a manageable first experiment could be a repeatable document review with a defined deliverable. Choose work where you can explain what is included, what makes a file unusually complex, and when a separate scope discussion becomes necessary.
Write those boundaries before comparing prices. Otherwise, a quick file and a difficult file can appear to be examples of the same service even though they require very different effort. Record the document volume, number of review rounds, urgency, and other factors your team considers material.
Choose a small initial group of completed matters as a baseline. Review the individual files as well as the average. A baseline dominated by easy matters can make the proposed price look comfortable while hiding the cost of exceptions.
Measure the whole delivery process
For each pilot matter, record total staff time by role, including setup, checking, corrections, and client follow-up. Keep a separate note of tool costs attributable to the pilot and any outside expenses. Record whether the work met the firm's existing review standard and whether it required avoidable rework.
Keep the measurement rules consistent between the baseline and the pilot. If the old process counted partner review but the new process omits it, the comparison will overstate the improvement. If the team is still learning the tool, identify that training period instead of silently folding it into a permanent assumption.
This is a proposed management exercise, not a claim that an AI tool will save a particular number of hours. The useful result might be faster delivery, fewer interruptions, or a finding that the tool adds more checking than it removes.
Use a simple example without mistaking it for profit
Hypothetical example: a firm charges a $2,000 flat fee for a defined service. One completed matter requires eight total staff hours; a later, comparable matter requires six. Fee revenue per recorded staff hour rises from $250 to about $333, assuming the full fee is earned and collected in both cases.
That calculation is only a rough efficiency indicator. It does not account for the mix of senior and junior staff, compensation, software, overhead, or expenses. If review shifts from a junior employee to a partner, fewer total hours may still mean a more expensive delivery process.
Build the cost comparison separately using the firm's own cost assumptions. Label estimates clearly and avoid presenting revenue per hour as net profit. Also ask whether the apparent improvement survives on the more demanding files in the pilot.
Decide what the released capacity is for
Before changing fees, choose what the firm intends to do with any time released. Options include serving additional suitable clients, reducing overtime, improving turnaround, or giving lawyers more time for substantive review. Each calls for a different success measure.
If the objective is additional work, check whether intake can actually supply it and whether another step will become the bottleneck. If the objective is a shorter working week, measure working hours rather than treating unchanged revenue as a failed experiment.
Then compare a few possible fee and scope arrangements against the same cost assumptions. Treat the result as a basis for a considered pricing decision, with the firm's usual engagement and professional-obligation review, rather than an automatic price change.
Where Clio and Keeney Analytics fit
For firms using Clio Manage, Keeney Analytics' Matter Analytics provides a place to review matters, recorded hours, and supporting financial and invoice detail. Use that evidence to investigate selected files while keeping the pilot's scope, review outcomes, and cost assumptions in a separate management worksheet.
Keeney Analytics explicitly distinguishes matter efficiency from true profitability. Do not assume a dashboard measures AI savings or includes every delivery cost. The comparison is only as useful as the underlying records and the additional context your team supplies.
Set a review date and an owner for the pilot. At that meeting, decide whether to extend it, revise the scope, or stop. A useful pricing experiment ends with a documented decision about a specific service—and evidence the firm can explain to itself and its clients.