There is a great hue and cry these days heralding the replacement of the billable hour model with alternative fee arrangements (AFAs). The notion seems to be that using AI won’t change much about the practice of law and how we bill for services: we can just switch from the billable hour to flat fees and nothing will be all that different. That’s just flat wrong and avoids the real issue.
AFAs
AFAs can take the form of flat fees for the handling of a legal matter in whole or in part, subscription fees where a client pays one amount for a defined set of legal services over a specific time period, value-based fees, and similar arrangements. AFAs are designed to get away from paying for legal services purely based on the time spent handling the matter.
I pioneered the use of flat fees for serial litigation (similar litigation involving the same product in multiple jurisdictions) back in the 90s and learned a lot about how to make them work and their pitfalls and benefits. At the time, we entered the arrangement for several reasons that were unique to the litigation we were handling. But I fear that the present rush to AFAs may be for the wrong reasons and based on misguided expectations that will lead to disappointment.
Why Now?
AFAs are great in many ways. They lead to innovative ways of approaching the client problems. They lead to efficiencies. They lead to predictability. But some lawyers and law firms are fearful that AI will replace many of the jobs that they have traditionally been able to bill for. Worse yet, they see AI as a threat to the economically lucrative leverage model that has powered unfathomable riches particularly among Biglaw. Instead of thinking through what AI could mean for the handling of legal services they jump to the conclusion that AFAs will enable them to somehow continue what they do and how they do it without any economic disruption. That is not the case.
We can see the problem when we consider the billable hour model and why it has succeeded.
The Billable Hour Model
The billable hour model has been roundly thought of as one of those “can’t live with it, can’t live without it” things. Most outside and in-house counsel condemn it but so far few have moved off it.
There are lots of reasons but one that is particularly relevant as lawyers look to AFAs. Simply put, the billable hour model incentivizes thoroughness and quality. And traditionally it is that thoroughness and quality that leads to more successful legal outcomes.
The model ensures associates will research long and hard, for example. That in-depth depositions will be taken. That all-encompassing motions will be filed. That drafts will be polished and repolished until they sing. It encourages no stone be left unturned. The more time spent, the more likely the better result.
Where the model gets into trouble though is where the value of the work done is far exceeded by the value of the thoroughness. Spending $100,000 in time on a $25,000 matter leads to anger and frustration. And, yes, I know, the model doesn’t always yield a better result but on an overall big number basis, it’s hard to quarrel with the idea that, on balance, the more time spent, the better the result. And I’m aware that there is a point of diminishing returns that the model often exceeds. But if you want to encourage thoroughness and quality, the model beats any other.
The model has also led to extraordinary profits as using leverage (having teams of lawyers working on a matter) enabled massive number of hours — and fees — to be accumulated.
Enter AI
But with AI, things have shifted. It’s now theoretically possible to achieve some level of thoroughness and even accuracy in a fraction of the time. The result? Less time spent. Less billable hours. I say theoretically of course because the tools still can make mistakes and must be checked. But, again, on balance, there is little question that AI will allow lawyers to achieve the same and perhaps better results for less time.
Motions can be prepared in a fraction of the time and mimic the author’s previous work. Deposition prep can be aided by an AI assistant who sits with you and supplies hints during the deposition. Case themes and arguments and even problem ones can be created and identified at the click of the mouse. All these not only reduce the time to get to a result, but they may also even lead to higher quality.
So AFAs Are the Answer, Right? Yes, But…
So, the knee-jerk reaction is to move to AFAs. Time doesn’t matter. We can get paid roughly the same amount for handling a matter as we did when we billed by the hour. We just package it as an AFA, and everyone is happy.
But that doesn’t necessarily hold true. The ethical rules are pretty clear that you can’t charge for the time you would have spent doing a task that an LLM does in seconds. Merely repackaging the fee as an AFA but charging roughly the same amount doesn’t make that “reasonable” within Model Rule 1.5 of the ABA Rules of Professional Conduct which governs fees. Indeed, one of the factors to consider when assessing reasonableness under the rule is the time spent.
There is also the rule of unintended consequences. The key to making AFAs profitable is to do the work required for less time and cost than the fee itself. This means that any time and cost that can be reduced or eliminated leads to more profits. So, what happens when lawyers eliminate cite checking as a time savings tool and court sanctions result? What happens to quality when corners are cut?
And maybe most importantly, what happens to client service? One of the fundamental cornerstones of legal service involves the relationship between a client and their lawyer. Building that relationship and trust takes time. Time is often considered the enemy in many AFAs relationships.
I have seen flat fees do exactly that. In the early 2000s, I witnessed a company refer all its product liability work to a national firm for a yearly fee. The lawyers working on the files immediately began doing less. The client thought it could ask for more work from its lawyers. The lawyers came to resent being asked to do things. The client began to think it better demand more so it can get what it really needed. The arrangement bred distrust. No one was happy. It was abandoned mid-year.
Rethinking AFAs
To work, the underpinning notion of what AFAs can do needs to be rethought. First, AFAs aren’t a replacement for the fees generated by the billable hour model. Instead, the value of the AFA has less to do with time and more to do with profitability. For the law firm, how much profit can be made with a flat fee set at a particular point given the costs associated with providing the service is the key question. And how can it best provide the service given that cost level?
It’s more about profit than revenue. And profit not revenue or time spent has to become the metric that determines the value and worth of the AFA to the firm. It has to be less about cutting corners and more about accurately assessing the cost of what you can provide. And then accurately communicating the service you will be providing for that fee.
When I set up the flat fee for national litigation in the 90s, the interest of the client predictability, capped fees, and the need to move cases to mediation quickly lend itself to a flat fee that rewarded achieving these objectives. This is why it was successful. The lesson is that moving to an AFA must be to achieve well-defined objectives, not maintain revenue at the level achieved by the billable hour model. AFAs are not a panacea.
It’s a New World
As a profession, we need to accept that AI may be driving us to new realities. We will need to do things differently. We will need to deliver services and charge for them in new ways. We need to define and measure profitability in alternative ways. AFAs are great but they don’t mean we can just keep doing what we are doing and calling it a different name.
It’s not just changing horses. It’s moving from the horse to the car. In midstream.
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 AFAs Are Great — But They Won’t Stop AI Disruption appeared first on Above the Law.