Is AI and the billable hour an efficiency powerhouse?
- unlatchedlaw
- Mar 8
- 4 min read
This article is a partial response to the Thomson Reuters article titled “Pricing AI-driven legal services: The billable hour is dead, long live the billable hour.” by Bruce MacEwan and Janet Stanton
Client’s are increasingly becoming adversarial when it comes to the final bill. The billable hour is by far and large the dominant model for calculating legal fees, and is based on recording 6 or 15 minute intervals as separate items in a timesheet, with explanations as to what work was carried out for each item. Those bills are then refined and reviewed, with items written off for various reasons (most often that a client will not want to pay for excessive hours of a singular task such as drafting).
Despite years of analysts predicting the end of times for the billable hour, it continues to remain the king of the hill of remuneration models, and is likely to continue being so for decades to come. This is due to one fundamental reason – the billable hour model is predicated on the idea that the time spent on any given matter, regardless of the quality or utility of the time spend on said matter, is chargeable. In other words, it is a profit machine for law firms who inevitably will do far more than strictly legal work from the start of a case matter to its close.
MacEwan and Stanton state the following as a key criticism of the model:
“It prices all hours interchangeably. Should days of scut work reviewing flagged discovery documents cost the client many multiples more than your one brilliant flash of game-changing insight that took hardly an instant on the clock? Of course not, but that’s the reality we’ve created and still live in.”
Of the many diatribes of the billable hour, this is perhaps one of the worst. Firstly, the ‘days of scut work’ which lawyers must undergo during due diligence or discovery processes is a necessary part of the preparation. It ensures that both the client and the lawyer have their backside covered. We cannot, afterall, accept that clients have recourse through a negligence claim for seriously defective work, and simultaneously expect lawyers to base their advice on flashes of ‘game changing insight’ which are not reviewed multiple times before being sent out. Secondly, the argument is devoid of logic because, typically, a ‘brilliant flash’ typically comes following hours of gruelling thought and consideration. The thousands of documents contained in a virtual data room, for example, are poured over to help formulate negotiation and risk management strategies, which may boil down to one or two paragraphs. Were those strategies based on the few ‘eureka’ moments that a lawyer has, clients could be sure to have far more expensive and unpleasant legal encounters when loopholes – which could have been plugged with a few more hours of care – emerge.
The strongest argument which MacEwan and Stanton put forward in their article is the first key criticism which they note regarding billable hours:
“Hourly billing misaligns law firm and client interests. The client presumably is seeking a fast and reasonably priced response, but the billable hour works against both those results.”
There is no doubt that this rings true when considering what time sheets entries are typically populated with. ‘Attending emails’, ‘reviewing draft statements’, and ‘preliminary meetings on strategy’ are all examples of vague descriptions that have appeared on final bills. Other criticisms include transparency in timekeeping, minor tasks being billed in large increments, and most importantly ‘time padding’ by adding unnecessary work to inflate billable hours.
Hogan Lovells partner Sebastian Lach told the Financial Times that a significant shift can be seen in the business model of firms, as the focus turns to technology cost instead of hourly rates. However, does this require firms to rethink the model or combine the development of technology to provide the optimal version of hourly billing? Under this version, generative artificial intelligence (Gen AI) tools can reduce the heaps of grunt work that lawyers lose revenue conducting. This efficiency will not only increase the amount of critical human input needed, but should raise the average quality of due diligence, contract drafting, arguments before court, and many other skills which will only be developed further as Gen AI comes to the fore. Firms such as Hogan Lovells with Elevate, A&O Shearman with Harvey, and Linklaters with Laila are all brilliant examples of using Gen AI to revolutionise drafting, negotiating, due diligence, and many other routine tasks. But when will Gen AI change time keeping?
The optimal version of the billable hour model, therefore, will only satisfy clients to the extent that technological development allows for extremely more detailed and accurate time entries in significantly shorter time frames. This element is what any current time tracking application or software is missing. Consider ability to, for example, speak into a microphone and provide a sizeable analysis as to what was done in any given 9 hour period, and have the software review and refine that voice message to produce an accurate time sheet with all elements of the project properly itemised and with succinct summary for clients. If Gen AI could ever rise to this fantastical reality of timekeeping, all of the aforementioned clients’ concerns could be culled, whilst providing efficiency gains for both lawyers and clients alike.
Despite the reign of the billable hour, fixed and contingency fee models have been pulled out of retirement as firms seek to compete for business, and clients set increasingly demanding thresholds for their willingness to pay. A survey of fee earners in 2021 found that 43% of UK law firms with more than 100 lawyers offered alternative fees like fixed or capped fees. Yet law firms continue to bill more hours and charge higher fees, increasing their yearly profits whilst clients seek lower costs.
The truth is, the labour of theory of value, which bases the value of any given commodity on the labour required to produce it, will always triumph in the legal world. Clients need lawyers and until Gen AI forces maximum transparency and accuracy in the work that is itemised in a timesheet, we are a long way away from utility (as opposed to labour) determining the value of legal work.
Comments