At a recent conference at which I spoke, a member of the audience raised her hand and asked why law firms are having so much trouble determining the ROI of artificial intelligence tools that they are buying. I thought about it for a minute, and then it hit me: it’s because they aren’t getting an ROI in the traditional sense. If law firms are using AI as they should be, it’s their clients who are getting the ROI.
Here is what I meant. If law firms bill by the hour and if AI does what every vendor says it does, its use would, by definition, reduce billable hours. That, in turn, reduces revenue and profit. It means, instead of a return on investment, they are actually taking a loss on investment. Call it the LOI.
(There is an important exception. When AI tools perform back-office functions, the opposite is true. These tools can reduce administrative costs, eliminate non-billable time, speed collections, and allow fewer people to perform internal functions. And all that reduces costs and increases profitability. That ROI can be and probably is being measured.)
For substantive tasks, use of AI will undoubtedly replace work done by associates and paralegals that was billed by the hour. It’s a net loss. No wonder firms are reluctant to use AI tools. No wonder they can’t figure out the ROI. There’s not all that much.
A Paradox
It’s a paradox: when AI is used by law firms, there is indeed a calculable ROI for the clients that are charged by the hour. If their lawyers use AI tools, the client’s bills simply will go down and go down by measurable amounts. When AI turns 15 hours into one, the clients are happy.
So, clients are all too eager for firms to make investments, often substantial investments, in AI. But these investments and AI use only help the clients and hurt the lawyers that just spent a fortune on technology. No wonder everybody keeps talking about ROI at conferences without quite managing to find it. It’s a difficult and, to be frank, somewhat unfair situation.
Another Phenomenon at Work
There’s another phenomenon at work here as well. Traditionally, under the billable hour model, clients would demand that work be pushed down to the lowest-cost provider while maintaining the approximate level of quality. It might take more time to do the work, but because the rate was lower, the net cost to the client was less. We’ve all had the conversation: why can’t you get an associate or paralegal to do that? I don’t need to pay you at your rate for it.
Hence, law firms built leverage pyramids. Those pyramids enabled firms to satisfy the client but still make money. And any loss of revenue from the work done at the reduced rates was made up by forcing associates to work long and hard to keep the machine running. Indeed, firms not only maintained profitability; they often increased it.
But with AI, that model is no longer as effective. AI eats the pyramid from the bottom up. It can do a lot of work that was previously pushed down to humans. Faster and at almost no hourly cost to the client. The profitability protection the leverage model provided essentially evaporates. Which is another reason firms aren’t itching to drive the shiny new car that they were convinced by vendors they absolutely had to have. It just sits in the garage, unused, and frankly, unwanted.
Is There a Solution?
One typical solution offered to the paradox is for firms to take a long view and say to themselves that if they reduce their bills by using AI, they will get more clients in the future. The problem is that’s a pretty indirect ROI, and it’s difficult to put a number on. It’s little more than a hope. And hope doesn’t sell well to a room full of angry equity partners wondering why their end-of-the-year distribution went down.
Another would-be solution is for lawyers to just up their rates to more than make up for the overhead cost of the AI tools and the lost hours. But to make up that profitability gap, rates would have to rise to astronomical levels. And clients would balk.
Do you really believe a client would willingly pay $5,000 per hour? $6,000? $10,000? I practiced for a long time, and I can assure you I never met a client whose response to a $10,000 rate would be wonderful, where do I sign?
Then there’s the favorite answer by pundits. Move away from the billable hour and to an alternative billing model. Flat fees. Subscription-based billing. But culturally, that’s very hard since everything about the modern law firm revolves around the billable hour. Compensation. Advancement. Partner status, power, and prestige. It’s the operating system.
And even if firms move away from the billable hour, it’s hard to believe that they can maintain the same level of profitability. Having done flat fees, I know how the client thinks: they are all for flat fees, provided the fee is less than what it would otherwise be. And it’s also challenging for firms to reduce the costs of producing the work enough to make up the difference.
It’s also difficult to calculate ROI based on that type of move. There’s an incredible amount of disruption to changing operating systems, and to be honest, some clients even resist such a move, thinking that if they can force their firms to use AI, the billable hours will be reduced, as will their bills. So why move to a different sort of fee structure that would not have the same bang for the buck?
Can We Talk?
But the best solution is something law firms and clients are reluctant to do. Law firms and clients need to talk to each other. Duh. They need to have honest conversations about how work is done and what it costs. I say reluctantly based on experience. Every time management or a well-heeled consultant tried to sell some “talk to your clients” initiative, that’s where it ended. With just talk.
Lawyers are an independent breed. They don’t like being told how to communicate with their clients. They don’t like having hard conversations about these subjects, particularly when it comes to money.
And make no mistake, for these kinds of discussions to be productive, a few things have to happen. Law firms, on the one hand, would have to realize that they have a duty to use tools that allow them to represent their clients effectively and efficiently. Clients, on the other hand, would need to understand that while they are entitled to the benefits of technology, firms still need to make a profit.
They can’t do good work and retain top-of-the-line talent unless they can adequately compensate those doing the work and keep the lights on. Clients also need to recognize that forcing firms to make substantial AI investments which strangle their profitability is not sustainable.
And clients need to make genuine commitments to their law firms that if they employ AI, the firms will get more work. Not just mouth it with what feels like a wink and a nod. It can’t be like the discount spiel I used to hear from clients: if you discount your rates, we will send you more cases. Somehow, all I got was less money for the same amount of work.
In short, clients need some ownership in how their firms use the technology. Firms and clients need to design workflows together. Both sides need to agree on what efficiency looks like and who receives what portion of the resulting value. Clients and their lawyers need to sit down together and determine what work AI can improve, how much investment is required, and how the benefits and costs can be shared. Not just realized by one side of the relationship.
Right now, what we have is a standoff: the client’s AI ROI is often the law firm’s AI loss.
So, can we talk?
Stephen Embry is a lawyer, speaker, blogger, and writer. He publishes TechLaw Crossroads, a blog devoted to the examination of the tension between technology, the law, and the practice of law.
The post AI ROI For Law Firms: There’s Not Any appeared first on Above the Law.
Recent Comments