This is how the billable hour dies

Let me tell you a story about the AI-driven evolution of pricing in the legal market. It might not happen for many years. It might happen much sooner. But when it does, I expect it'll look like this.

Illustration by Midjourney

There are three things to know about hourly billing in the legal services marketplace.

1. The Well-Known Truth: Lawyers love it. Hourly billing is a simple, convenient, risk-free, and incredibly profitable pricing mechanism. Lawyers have been happily using it for decades, and most of them see no reason why they should ever stop (aside from their own well-being, of course).

2. The Less-Publicized Fact: Law firms need it. They don’t know any plausible way to price their work other than by measuring and charging clients for the time and effort required to generate it. And anyway, it’s too deeply entrenched in the firm’s culture and infrastructure to change now.

3. The Dirty Little Secret: Many clients don’t actually mind it. In-house lawyers grew up on the billable hour, they know how it works, and they know how to manipulate it through discounts and write-offs. Less experienced clients don’t like it, but they don’t entirely trust fixed fees either, so better the devil you know.

That’s why hourly billing has survived so long and dominated the legal market (with a generous assist to lawyers’ regulatory monopoly). Vanishingly few law firms have ever given up on hourly billing and moved to alternative fee structures — it’s rare to find a firm that wants to do it, and it’s ultra-rare to find one that could actually make the transition in practical terms.

The safest bet in the world is that hourly billing will continue to flourish in law firms, day over week over month over year. I myself will confidently assert that next year at this time, hourly billing will still be the dominant pricing and value measurement methodology in law firms.

Now, I’m going to tell you a story about how everything that I’ve just described will come to an end, thanks largely to Generative AI. Your law firm, today, is almost certainly not interested in or capable of developing new pricing and productivity metrics that aren’t time-based. This story is for the day that you are.

There’s a rapidly growing mountain of evidence that Generative AI is, or shortly will be, capable of performing many written, analytical, and creative tasks now carried out by legal professionals. There’s also good reason to believe — based on factors such as client demand, competitive pressures, and lawyers’ own desire to use generative AI to eliminate boring and laborious tasks — that law firms, whether they want to or not, will feel obliged to experiment with, and eventually adopt, Generative AI into their workflow.

Lawyers spend many hours every day performing myriad billable tasks for clients. When Generative AI is applied to these tasks, however, the amount of time and effort required to complete them decreases — first slightly, then significantly. This process is just beginning, and I haven’t met anyone who thinks it will stop or reverse.

Generative AI will not eat all the hours. There will still be traditional hourly work for lawyers, quite a lot at the start — some work will not be amenable to Gen AI involvement, although other types will be better suited to more familiar technologies like automation. There will also be some lawyer work involved in prompting the AI, assessing its output, fixing its errors, and turning drafts into documents and eventually into client deliverables.

But as the AI improves, the prompting and assessing will take less and less time, and the AI’s output will become closer to being client-ready. Within a few years, depending on the nature of the tasks and how complex they are (or aren’t), what once occupied 100 hours of a lawyer’s time will take closer to 90, or maybe 70, or even 50.

What will happen then? The lawyer will seek more traditional hourly work, of course, because they need those hours just to stay afloat or to remain employed at the firm. Maybe the lawyer will find and take on a new client with work of this type, thereby filling in some of the space in their schedule that the AI inconsiderately “freed up.”

But since this work also will be amenable to the application of Generative AI, the lawyer will need to keep increasing the number of new matters just to get back to their pre-AI workload. That, in turn, will increase the amount of time and effort required for business development and client management, none of it billable.

Yet despite this intensified effort, the lawyer’s overall “productivity” will stay the same as it was before — at least as far as the firm is concerned, because in law firms, “productivity” is defined in terms of the number of hours billed. Frenetically working and billing the same hours, but for more clients and with more rainmaking and relationship responsibilities, the lawyer is on a treadmill, one that keeps speeding up.

The lawyer might not like this. They’re working harder than ever before for the same reward. Not only that, but the work’s not getting more interesting, and the AI is growing more capable of doing more of it all the time. The lawyer will start looking for better options.

One thing the lawyer might do is approach the client and ask to perform higher-value work. This would be work that’s more creative, strategic, forward-looking, and risk-assessing than the old legal work that Gen AI has significantly downsized. This work, the lawyer explains to the client, would be more valuable and can therefore be charged at a higher rate than the basic tasks. In this regard, the lawyer will now find Generative AI to be an ally, because it’s extremely creative and can spark and boost the generation of new problem-solving ideas and perspectives.

If the lawyer succeeds, then they will fill in some of their freed-up time with more sophisticated work that not only is more interesting than what they were doing before, but also can be billed out at a higher rate. Terrific!

But there are still a few problems. First of all, it will take some time to persuade the average client that they should pay a lawyer to do strategic risk assessment and creative opportunity identification. Because while that kind of thing sounds great, it isn’t nearly as vital to most clients as the deal that must close tomorrow or the lawsuit that must be defended or settled the day after. Clients haven’t proven famously receptive to pitches along the lines of, “Here’s something you never asked me for, and you can pay me a lot of money to do it.”

Not only that, but while the work might be more interesting, it usually won’t require the same volume of time and effort that traditional work did. Higher-value work doesn’t necessarily mean higher-hours work; often, it can mean the opposite. Lawyers frequently say things like, “I finally solved the problem while taking a shower.” It’s difficult to bill showers by the hour.

The lawyer increasingly starts to feel shortchanged. Here they are, coming up with great ideas and insights that could save the client hundreds of thousands or even millions of dollars, but all they get is several hundred bucks an hour. The firm won’t make up the difference — it only counts the hours that the lawyer bills and doesn’t much care about the value provided to the client. The value of these higher-level services simply isn’t measured well by the time and effort it took the lawyer to deliver them.

Meanwhile, the law firm is going through a similar process. It’s having trouble finding more hours for less senior lawyers to fill and bill. The senior partners aren’t interested in helping — they’re seeing their own hours start to shrink, as Generative AI steadily climbs the value ladder and starts taking chunks of their work. So they hoard their hours even more than before, leaving junior partners and associates to sink or swim on their own — further compromising the firm’s leverage.

The firm begins to feel profitability pressure. It reduces its overhead costs by cutting office space and letting go of idle associates and unproductive partners, but there are hard limits to how low those costs can go. Not only that, but client demand has softened across the board, as clients discovered that using Generative AI can help them complete some of their own basic work in-house. This prompts clients to pull back other basic tasks, and even some intermediate ones, from their outside counsel.

The firm watches as its annual inventory of billable hours keeps shrinking, while the business development costs of constantly seeking more hours rise. Rate increases for higher-value work can’t fill the gap — clients keep reminding the firm that the introduction of all this new technology was supposed to make legal services more efficient, not more expensive.

Eventually, the firm finds itself facing a conclusion that it’s tried very hard to avoid: Selling hours simply isn’t as profitable as it used to be. As the firm becomes more engaged in complex, high-value tasks, lawyers complain more often that the time-and-effort-based pricing system doesn’t serve their interests. They can’t raise their rates to $1,000 or $2,000 per hour, but they don’t know how else to get paid what they think they’re worth.

So now both lawyers and the firm are looking for alternatives. Of course, people have been talking about fixed fees since the 1990s — but going back to our three facts at the start of this post, hardly anyone has used that method. But now hours are fewer and less reflective of the value generated. Might it be time to reconsider?

This would not be an easy path. Lawyers remain deeply attached, financially and culturally, to the billable hour model, and change has always been difficult for members of this profession. Clients, who’ve finally seen the worm turn and are now actually reaping the benefits of hourly-based billing, need to hear very persuasive arguments that project-, matter-, or client-based billing serves their interests. But there may come a time when the firm and its lawyers feel they don’t have much choice except to try.

And that, I expect, is how the billable hour will finally go into that good night. Not because some dork with a blog thought it would be a good idea, but because the nature of legal work changed in ways it never had before. Technology automated and augmented legal tasks so thoroughly that the value of the work radically diverged from the amount of time required to do it. The base rationale of hourly billing — that it was an imperfect but serviceable proxy for the value of legal services — couldn’t be maintained anymore. Selling legal services by the hour became too unprofitable.

I have no idea when this will happen. It won’t be this year — your firm doesn’t have to place this on top of its strategic to-do list. But if there was ever an issue that qualified as “important but not urgent,” this is it. Every day you wait to get started is another day lost, someday down the road, to competitors that began dealing with this sooner and faster than you did.

So assemble some of your most creative, forward-thinking people, and ask them: “If the firm could no longer bill our work by the hour, how could we turn a profit?” Give them this article from 2012 to get them started. Show them the firm’s financials for the last 24 months, so that they know how you’re making money now. Have them speak with clients, technology experts, and pricing consultants for insights — might as well get them to ask ChatGPT, too. The answers you get will form the basis of your future strategic plans.

There’s actually a fourth thing to know about hourly billing. It comes from 20th-century American economist Herbert Stein, and you might have already heard it: "If something cannot go on forever, it will stop.”