The legal world has debated the demise of the billable hour for years, with generative artificial intelligence merely the latest rationale offered for its extinction. The traditional billing model persists for many reasons, although law firms are moving toward new models as in-house counsel realize efficiencies from legal technology and resist certain fees.
Legal Dive spoke with Rhys Hodkinson, chief revenue officer at Definely, which makes software for drafting and reviewing legal documents.
Hodkinson, an attorney in corporate law before he joined London-based Definely in 2020, spoke about how lawyers consider the value of legal work, why a “self-serve” model is becoming attractive for both in-house teams and their outside firms and how new tools will filter through the profession in coming years.
Editor’s note: This interview has been edited for clarity and brevity.
LEGAL DIVE: When you speak with potential customers about your company’s software, are you finding more traction with law firms or in-house legal teams?
RHYS HODKINSON: In-house is starting to catch up quite quickly with law firms in terms of our pipeline. I think there are a few things that we’ve seen with respect to in-house. Some of the feedback we’ve gotten comes from a cost-containment perspective. Every in-house organization, not just the legal team, is trying to make sure they’re containing costs. But the other one is to create self-serve. The goal is to allow people within the organization, who might not be with the legal team, to come in and self-serve on the basics, freeing the legal team to do higher value tasks.
Let’s discuss billing-model disruption and how law firms consider that possibility. Do they want to know about their economic future with this AI technology?
It’s always the question. The hourly model creates a perverse incentive for inefficiency. But obviously law firms have been doing pretty well of late. It’s hard to convince millionaires there’s something wrong with the business model, right?
If you look on the transactional side, where law firms write a lot of their fees, it’s banking and finance, private equity, M&A. But if you look at the other advisors on these deals and how they charge, it’s quite different. With an investment bank and a law firm working on a transaction, the investment bank always charges a percentage of the total deal fee. On the total deal amount the investment bank fees end up being multiples higher than what the law firms typically end up charging on a per-hour basis or on a total basis. The lawyers are not the most expensive people necessarily on the transaction, even if on a per-unit basis they look like they are.
When you get a bill from a law firm, they literally tell you what they’ve done for every six-minute increment of time. More often than not, in-house teams will push back on that. That’s why you have about 80% realization rates in law firms, with 20% of the time getting written off because in-house is starting to get a sense of what they’re doing. If you’ve ever seen an invoice from McKinsey or Bain or BCG or Deloitte, it just says time worked is X and you pay us this amount of money. If you start taking away the billable hour, what do you get instead? Do lawyers start saying, “Hey, we’re going to charge them value.”
I’ve got some examples of transactions I can’t disclose, but the tax advice that was given saved hundreds of millions on a transaction by virtue of structuring it. Now, would you rather have that lawyer that took six hours, and write your check for six hours, or would you rather them take a percentage of the value that they’ve provided? So, I think you’ve got to be cautious. What’s already happening with the few quite forward-thinking law firms is the low-value tasks are very likely to get automated. There will be a lot of the kind of quick turns of reviews, access to basic tasks where you don’t need teams of lawyers charging per hour.
The firms that get smart … are almost offering a subscription model to their clients to basically automate those low-value tasks for them using fine-tuned large language models that they’ve trained over their data. Let’s say on an employment contract, I wanted to get my law firm to do a double check of it, but every single time, they charge me for two hours, and it’s exorbitantly expensive, and they’re going to come back with the same two points. If I had a subscription service, which would be quite expensive, like $20,000 a month, I can at least say this is what we’re going to spend. And I get these services. I can say, “Draft me an employment contract that deals with ABC,” and on the back end, you’ve got a proprietary software that’s being developed, probably by the law firm, with the help of external partners, that could be client facing, and it spits out a first draft based on your proprietary data set. That’s powerful. The law firm is still capturing revenue from that.
So that $20,000 is less for the firm, but a greater value to the GC?
You might have got $30,000 on these kinds of tasks, and say I’m just going to charge them $20,000. I can get 3x the business, and I might even be making a better margin on it because I don’t have a team of very expensive lawyers based in London or New York reviewing it each time. The other option is, well, we’re going to lose that revenue regardless because the in-house teams might start adopting the software to have a first go at those things, instead of coming to us. The smart way to do it is to understand that we can’t compete here, but at least we can start operating subscription services to help with those tasks, take it off people’s plate and capture the same amount of margin by being clever about how we do it.
In-house counsel may be frustrated about being billed for certain things they believe technology should handle now efficiently, and many law firms know they’ll need to adapt. How do the parties negotiate this?
BT Group in the UK [formerly British Telecom] is one of the few companies that has announced their law firm panels every four years. The main criteria as part of which law firms got on their panel this year was who is using technology to deliver legal services in the most efficient way? They have law firms come in and pitch on not only their experience but how they deliver legal technology in an efficient manner. The second thing that they did is they said the firm that's best able to show a use case of AI that helps them as a client automatically gets on to their next panel.That would be the in-house teams pulling the law firms along to say, “Hey, you want my work. You need to explain to me why I’m not paying for stuff I shouldn’t be paying for, and show me the tech you’re using to do that.” I think the law firms that don’t get ahead of that trend will be left behind.
I think everyone values the in-house counsel-law firm relationship. Where it’s become strained is over “I’m not paying for that. That is not what I asked you to do. I don’t need three junior associates on this simple question. Why did you put them on that?”
They are willing to pay for stuff, and there’s stuff that they aren’t willing to pay as much for. And the law firm’s job should be to say, what are firms pushing back on? Where is the work that they’re not comfortable paying these amounts for? Maybe we don't offer it and we have to run a leaner operation. Or the smart thing to do is, we create more of a mechanism whereby people can come in using large language models that are fine tuned around our proprietary data to be able to self-serve on a lot of these low-value tasks. So we protect downside risk on revenue. And what we shift as an organization is to say, as soon as you move away from anything that’s low-complexity, high-volume, and you want to start getting people involved, then you come to us. And your total spend with us might be the same, or might only go down a little bit, but you feel less like you’re paying for things that aren’t adding value.
Are companies beyond BT asking law firms these same efficiency questions?
Yeah, 100%. People on in-house teams are using things like ChatGPT or whatever to draft first versions of a contract, even if they’re rubbish. And they’re sending it to the law firm to say, “I don't need you to draft me a new contract. Just read this over and make sure that there are no mistakes in it.” They’re shifting from saying, “Can you do 80% of the work on this, and I’ll do the 20% review?” to “Look, I’ve done the 80%; can you do the 20% review?” And so law firms need to smarten up and say, “Fine. If they’re going to be doing that, then let us do it through a safer portal that’s trained on proprietary data that’s had lawyers look at its outputs and fine tune it to make sure it’s accurate.” And we can capture some of that revenue, because people are doing that already. It’s starting to happen. So law firms just need to wise up, otherwise they’re going to get left behind.
As AI becomes common and billing models are sorted, where will these technology tools sit? With the in-house teams, or do clients want these to be housed mainly with their outside counsel?
I think you’ll see law firms be much more use case-specific. You’ll get full-service law firms that would have insolvency units, litigation units, private equity units, M&A, structured finance, and there might be a need to have a real estate product that’s focused on just solving for real estate issues. They might need a specific product that’s helping on transactional drafting that’s using generative AI to solve for certain PE deals. For in-house teams, it’s always going to be hard to justify having 10 very bespoke solutions. I think you will find that the in-house teams will say, “Look, this is really useful for 60% of our work, and we can afford to pay that if we’re going to get the biggest ROI across the business for this.” But on these external models that are specific and require a lot of implementation, require a lot of effort to drive adoption, we will rely on our law firms to focus on those very specific areas.