Trial Lawyer View Podcast
Shim Hirsh

Jason Lazarus00:00:00

There was a firm I was talking to at New York. What’s your conversion rate? And he told me the number and I’m like, excuse me, that’s a percentage of what exactly is the percentage was pretty atrocious. It was below 50%. And I said, stop what we’re talking about. I’m about an hour and 25 minutes away from your office.

Shim Hirsh00:00:14

I’m coming. And we’re going to solve for this intake problem because that is a massively their leaving about $2.5 million of inventory value a year. And as a percentage of their revenue, that was about 25%. Shim Hirsh.

Jason Lazarus00:00:28

Founder of Better Works, a company that builds and manages AI powered workflow automations for personal injury law firms, helping them improve intake, case management and client follow up. He spent seven years in technology and operations at Morgan and Morgan, the largest personal injury firm in the United States, with over 1000 attorneys, 140 offices, and more than $2 billion in annual revenue.

Shim Hirsh00:00:51

Out of the 1000 phone calls were 950 of them calling about trying to buy a car insurance policy and 50 of them only calling but injured in a car accident. Right? Then you have a marketing problem, and if you sign those 50 at a 50, that’s, by the way, excellent intake. But if you have 950, those phone calls are actual accidents. People that were injured in auto accidents and are seeking representation and you only sign 50 of them, then that’s a problem.

Jason Lazarus00:01:11

In this episode, you’ll learn how to measure your firm’s true intake conversion rate, how to distinguish an intake problem from a marketing problem, and how to uncover the hidden revenue leaks that could be costing your firm millions each year. I’m Jason Lazarus and this is Trial Lawyer View. So being from Orlando, having grown up here and having witness Morgan calling in Gilbert, which is what it used to be to transition into Morgan and Morgan has been a crazy thing to watch. And you spent seven years inside of that machine, inside of Morgan and Morgan as a product leader responsible for many products, technologies, initiatives, reporting to the CTO and COO. So you are part of a team that ran technology and ops across more than 100,000 active cases, 140 offices, thousand plus turns, most personal injury firms and personal injury firm owners never see personal injury law at that scale. What did you see from that seat that you would like every person, firm owner, to see if they could for even just a week?

Shim Hirsh00:02:25

It’s a really great question. You know, so as I talked to a lot of firms now, what I realized is that the average 25 person plus minus firm experiences, the same bottlenecks and pain points that Morgan does. It’s just on a much larger scale, of course. And the real problem is not it’s not a law problem. It’s not a lack of legal skill. It’s an operations problem. It’s a data problem. And so a firm of Morgan size, they either live and breathe and operate on that data or they choke on it. And so I wish that any legal professional in the pie space would be able to really see what true transparency into your inventory looks like. I think Morgan really, really does that very well at a very good detail. And if you can see what transparency looks like, you start seeing a lot of answers to questions you would never even ask, right, just to answer themselves. And so that’s a really a really important part that I wish everyone can experience.

Jason Lazarus00:03:25

You know, it’s interesting that you say that because it seems more and more today, there is this focus of looking at your data if you are a law firm, instead of it being the practice of law. It’s a focus on the business side of law. And that data now is what those law firms run on instead of what it’s been, which is more by the CD or pants type of stuff. I’m curious with what you saw, you know, what broke your assumptions about how personal injury firms should run based upon what you saw sitting in that seat?

Shim Hirsh00:04:13

I think that my assumption is that as you scale right, it’s required to kind of operate on this data in order to continue to deliver service. You might think that a byproduct of that is a sacrifice on the actual service that you deliver. And I spent a lot of time on this at Morgan and Morgan, specifically the client experience. And what I learned is that operationalizing data to to really build and grow and scale a firm, you can actually start to deliver a better service, more consistent throughout, you know, all the different geographies that you operate in, even if they all have very, you know, different operating structures and nuances based on jurisdictions, etc., I think you can actually achieve both a better experience, a better service that you deliver and have that ability for for Hypergrowth.

Jason Lazarus00:05:06

If you notice how most conversations about law firm growth stop at intake marketing and trial results, but rarely addresses what happens after resolution, that’s the Synergy focuses on. We partner with personal injury law firms to manage the operational side of health care, lien resolution and compliance, brain structure and clarity. One of the most overlooked but costly parts of what your team does day in and day out. If you’re thinking more seriously about operations beyond the courtroom, you can explore Synergy Services through the link in the show notes or by scanning the QR code on the screen. Now back to today’s conversation. Seeing where I sit and having seen a lot of things in the growth of our business, the customer experience or client experience is such a critical thing. And in particular, I’ve talked about this with some others that have been on the podcast. Personal injury is intensely personal. Having been through myself, it is very difficult thing to to navigate. And even with a great lawyer, which I had a great personal injury lawyer handle my case and still it was tough. It’s just not. And having had all the experience I had in the personal injury realm, it was still tough. So I think that that’s such an important part of all of this is, you know, as you scale is figuring out how to make sure you consistently deliver on that excellent customer experience. And you know that that’s what continues to get you more referrals to your law firm. It’s a lifeblood.

Shim Hirsh00:06:44

I was about to say it’s free advertising, right? It’s only in your best interest as a firm owner to deliver the best service. I mean, let your customers be your advocates.

Jason Lazarus00:06:52

With technology being at the forefront of everybody’s minds and things that are going on with AI and AI being integrated into different software solutions. I’m curious, this idea of switching case management system platforms, you’re making a case against switching, and I’m curious about that, because it seems like whenever I talk to somebody that’s been on the podcast and they’re talking about their case management system for their firm, they’re usually have a fair bit of criticism of it and even with the infusion of AI. But so why? Why is it the wrong move for firms to even consider switching at this point? And what should they do instead?

Shim Hirsh00:07:44

Let’s zoom out a minute and understand. Or what a CMS, what software is supposed to be, what job is supposed to serve? I think it could really be broken down into two simple pieces. Number one, I think a CMS is supposed to be the source of truth for all your metadata, for your the incident information, important dates, client information, all the other parties, insurance coverage, all the stuff. It’s supposed to be the source of truth to hold all of that. And then number two, and this is the critical part, it’s supposed to be an interaction layer. So anyone in a firm can use this interaction layer to access the data that the CMS stores. Right. That’s the core of, I think, what a CMS is supposed to do. Most challenges I’ve seen with CMS is not in what it’s capable of. It’s really it’s the second piece. It’s that interaction layer, its users not seeing the right data or not capturing the right data, and not and not being able to action on, on those, on those things. And so, I mean, that’s why I think you, you start to have a lot of, a lot of shadow processes that kind of organically grow as a firm scales. Right. So let’s take an example. Small firm. When a case flips to lit right. There’s a pre pre file checklist. You have to make sure a bunch of pieces of information or an order before you can file a complaint and so on. Right. One person handles that. They get few a week, few a month, ten a month or whatever it is. And they can they can manage those ten pieces that they have to look at for the ten cases. And if there’s something they missed, they’ll they’ll make sure they get to it. As you start to scale and now you have a few people doing that 50 times, 60 times, 70 times, you realize that you have to start managing. All right, well, which one of these files got checked to make sure everything was, was was sound out of the ten pieces, which four are we missing? All right, so these four files, we have to go and get these few pieces of information. But there’s ten more that just hit my desk, and now I’m out on vacation for the next three days. Someone else has to pick it up. And so how do you solve for that kind of chaos? As you as you grow and the answer becomes, well, the spreadsheet. Right. And what is a spreadsheet really? If you break it down in this context especially, there’s just a very simple way for someone who’s non-technical and who’s not software engineer. It’s a very simple way for them to effectively create an interaction layer. Their effect. They have columns and rows now where they can track something happens on this and this row, did these ten columns get checked and so on, which is great. But what they also did with that realizing is created another source of truth. And that’s where you start to run into some problems, because now you have someone who doesn’t live in this mini team and they want to know, why is my case stuck in prelate in without being filed for three months and what’s going right? And that’s when you start to have problems that way. And so it sounds like, oh, it must be a new CMS. But that’s not the answer. That’s just a messy shadow process. And if you would work with it and maybe some some CMS, maybe you can’t work with. Right. We experienced a pretty wonderful in the industry, pretty wonderful cloud renaissance over the last 1020 years where everyone moved away from the on prem, old school Microsoft Access database. Things to like the cloud based software, whether it’s homegrown like Cleo or File Vine or a Salesforce solidify solution. Both great pros and cons on both sides. But generally you can you can shapeshift the interaction layer for your CMS in order to solve your operational needs. And if you think buying a new CMS is the answer, it’s like, okay, Jason, let’s say I came to your house and you gave me the grand tour, and I say, your bedroom is full of laundry and dirty socks and everything on the floor. Am I going to tell you? You know what, Jason? I think you need a new house. Because what’s going to happen week after you move it to a new house? You shit all over the floor, right? So so that’s that’s not really the answer. And not to mention the actual cost of moving platforms is not not small. And I don’t just mean the marginal cost of the new software or the cost to build it out or the cost to migrate. Those are all costs that you can forecast and even downtime, right, for your operations. People develop habits and skills over years of using a particular piece of software to switch that out. You’re asking your it’s a big ass to have someone get used to a new interface, and now you know they’re going to operate slower because they have to look for things in different places. It’s a it’s a big ask. And sometimes I know in some cases, firms can lose 10 to 15% of their staff during these massive migration processes. So essentially short answer switching CMS is is not always the the problem. I would urge folks to really think about what is the core problem you’re trying to solve. You know, maybe Jason needs to fold his clothes in the morning. I don’t know, maybe the core problem is it can be solved in a different way.

Jason Lazarus00:12:14

You bring up a really important point that I’ve been talking a fair bit to the firms that we work with about, which is this idea that, you know, firms will tend to keep relic processes or manual processes, for example, in our world, doing the identification and verification of health care recovery obligations in most law firms today, it’s done with spreadsheets. It’s done with processes that sit in people’s heads that they leave it gets goes with them. There is no standardized process. There’s no single source of truth when it comes to this kind of data. They have no way of really getting that data into their system. So in you know, in my mind, that is such a huge hindrance to scale because if you don’t solve those issues, then as you try to scale, things break and you’re depending on humans, when at that moment the most critical thing is the dependency of the data, the being able to depend on the data that’s there. And so that that point about the single source of truth. And in my mind, the key thing today is deploying solutions that maybe are outside CMS, but that integrate with that CMS that are built to flow data into the CMS, whether that’s customized AI or, you know, niche AI for specific workflows that today are are, you know, currently not dealt with in a technologically advanced way. Those are the opportunities to me for real change and advance to help law firm scale.

Shim Hirsh00:14:14

Yeah, for sure. I mean, to go back to the to the earlier point, I was talking about of transparency, really of if you had real transparency, the questions would start answering themselves. I mean, my super simple example of that is, I think what you’re talking about as far as the source of truth in the right, in the same place, say, let’s say intake is the easiest thing to understand, right? Say a firm gets a thousand raw phone calls a month and they get 50 retainers at the end of it. It’s had a good outcome or a bad outcome. Answers you have no clue. Right? Out of the 1000, phone calls were 950 of them calling about buying, trying to buy a car insurance policy and 50 of them only calling but injured in a car accident. Right? Then you have a marketing problem. And if you sign those 50 out of 50, that’s, by the way, excellent intake. I would say that’s perfect. 50 out of 50, right. But if you have 950 of those phone calls are actual or actual accidents, that people that were injured in auto accidents and are seeking representation, and you only signed 50 of them then, right? That’s that’s a problem. And if you had a separate spreadsheet that tracked, okay, well, how many calls had to answer? How many calls did I how many retainers did I send out? How many times did I fall if that didn’t live in the same place? How are you going to know how to optimize? And if you add if you spend all the time and money to generate 2000 phone calls and you now you only had 100 fans, right? That’s not. And so keeping the right types of data inside in one place that you can easily get a bird’s eye view and drill down to the details. The critical part.

Jason Lazarus00:15:43

Yeah, it’s it’s interesting. We’re a Microsoft shop. And so, you know, we use power BI, which pulls, you know, reports out of the data that’s in our Microsoft Dynamics platform. And it’s amazing. We see with that and what you can do with it. And build customized reports and drill down into the data to give you a much better view of really, the performance of your business in any aspect of it. So to your point that that’s such a beautiful thing that you get by virtue of having all of that data resident in one place, but it’s amazing how how many times that that’s just simply not the case in our little world. I think, you know, in the personal injury space, it’s it’s more of a movement now towards that. But, you know, historically it just has not been. I’m curious from your view in terms of workflows, having seen what you you, you saw Morgan, Morgan, you know, as a firm scales and grows what tends to break first is it intake case opening record management, the flipping to litigation process. And what what is it that that breaks and what starts to give you that early warning. Hey, this is going to be a real problem. We don’t solve it as we continue to scale.

Shim Hirsh00:17:20

The answer is it all breaks, right? That’s the answer breaks everywhere. Some firms are, you know, laser focused on intake. Some firms laser focus on, you know, flipping to lit. But it all breaks when you start to really scale. I mean, you know this, right? In personal injury law, any sort of bottleneck or broken process, you only feel the pain later on its downstream where you start to feel the pain. Right? Like you were saying earlier, if you don’t identify the all the proper lien holders and the amounts early on, then by the time you settle now you’re chasing and running around trying to find, you know, accounts and invoices and things from years ago. It’s it’s it’s painful and there’s a lot of lost productivity. And certainly in that case, when everyone thinks it’s over, it’s a client experience problem. But I would say that in early warning sign is very simple. I would say walk over to any person inside of a firm who actually owns a caseload, who manages a caseload, who’s responsible for her caseload. Ask them to pull up a random one of their files, just scroll and click and ask them, where is this case at right now and what needs to be done next, and see how long it takes for them to answer. If you’re just looking at a blank screen with some simple dates and a name, and you see that the person is is recalling from memory, that’s usually a sign where, right? I mean, and my personal experience, I see this a lot in post intake, right. So intake is very easy to measure in all shapes and sizes like we were describing earlier. And so people are there’s also a tremendous amount of software ecosystem around intake. So people are really pretty solid. I didn’t take for the most part. I saw one terrible example, but for the most part, people are really good at intake. It’s after you sign a case. Now you have this like opaque process around, okay, we have to gather a bunch of information and get the client treating and right. So that’s like a gray area of what am I trying to capture? What are all the items? How long does it take to get each one? What does that unlock downstream? At what point during a trading phase and might start moving, start being more active, proactive on a case, starting to negotiate already, like those are the gray areas that lots of firms that I see people are making individual decisions on. And so you’re really at the whim of the person handling a specific case. That’s not a way to scale business.

Jason Lazarus00:19:40

So yeah, I’m curious about your experience in terms of operational breakdowns and the worst that you have seen and what it actually costs that firm.

Shim Hirsh00:19:56

Let’s see. So I can think of two examples. I can think of one intake example and a non intake example. So in intake example there was a firm I was talking to at New York. We were trying to figure out ways to shave days on desk during the litigation process because of course everyone knows that’s a lengthy one. And we were talking about the business in general, because, you know, it’s important to understand the context of the overall firm when trying to figure out even just one solution, our ones Pacific problem. And so we were talking intake. I was like, oh, what’s your conversion rate? You know, like how do you think about it? What is it? And he told me the number and I’m like, excuse me. That’s a percentage of of what exactly is it going. Yeah. The, the of the of the caller I don’t want to. The percentage was pretty atrocious. It was below 50%. And I said stop what we’re talking about. I’m about an hour and 25 minutes away from your office. That’s where I live. I’m coming. And we’re going to solve for this intake problem because that is a massive leak. I mean, they were leaving. It was wasn’t such a large firm, and they were leaving about $2.5 million of inventory value a year. And as a percentage of their revenue, that was about 25%. Right. And so that’s a lot of leakage for a pretty easy problem to solve. In another case, it was back to that case opening. Right. I call it intake 2.0 because that’s how I think it should be thought of. It’s of a of a structured process where you’re collecting police reports and liability, you’re collecting coverage and deck pages and trying to get amounts. Right. You’re really you’re collecting medical providers. It could be very structured if you make it that way. So rather than just like the opening of a case, let’s call it intake 2.0 and treat it that way. There was a firm that was it was taking them about 120 days or so to kind of get a full picture of the case until they called it in treating or medical management from calls of something else. And that was 120 days of just waste of time, not pushing hard enough to get police reports and building a process around it, not building a process around trying to get even first party coverage. Forget the third party coverage, just not building processes around understanding what these pieces of information are and what we can do with them once we have them. And so still working with that number, trying to solve for it. But that’s just an example of months of wasted time, wasted productivity.

Jason Lazarus00:22:23

You brought up intake. So I wanted to segway to that because from doing my research, I seen you say that most firms measure conversion wrong when it comes to intake. What is the reframe of that, and how does it change the way a firm should start and invest?

Shim Hirsh00:22:42

It’s exactly like we were talking about earlier about that, that case of 1000 calls and 50 retainers. If you truly understood where the breaking points are, you know what to solve for. So if you, for example, go back to that case of the thousand calls, if you if you if you saw that 950 were non-qualified callers, they were just calling about other things, then you call up your CMO and start asking some very serious questions about their ad copy. Right. Or but if but if it’s really just a conversion problem, then, okay, what are the documented steps that your that your pipeline or intake pipeline is taking. Right. So is it quick qualification of the you know, Sol injury jurisdiction case type. Right. All the all the classic criteria questions. What happens afterwards. Are you. Do you send a retainer on the same call. Do you do you send it up. You follow up after the fact. Do you send someone to their house in some instances, or mail something and send something in the mail? How much time is every minute? I mean, intake is really measured in seconds, right? So every second you don’t pick up a web form and call and call them is another second that they’re very likely to call the next, you know, turning on the list. And so if you go down the step by step and you find where the drop off is, you have to you solve for it. Maybe your send retainer process takes too long. It’s ten clicks and you do it post call and someone forgot about you and they call someone else already, right? I mean, understanding where the where the drop offs are, right? That that transparency that into the inventory to the process. Specifically here we’re talking about intake. That’s how you solve. Maybe you had too many calls in the in the queue on hold and you’re bending rate is whatever 1,520%. I mean that’s a crazy leak. So then hire more people to answer the phone. I mean, every data point has, it leads to a different answer, a different solution. And lots of times it’s just a tension, right? It’s not like, oh, let’s throw AI at the problem and hope that it can be solved. It’s just focus on where the core, where the actual root cause of the problem is and solver that not something else.

Jason Lazarus00:24:50

So in the end, let the data tell you where the issue is and then solve for that particular. Pretty simple. Assuming you you, you’re looking at the right data and have the right place and are able to to see it, which is, you know, to your point, such a critical thing that we’re talking about.

Shim Hirsh00:25:12

Here, I think it’s a discipline start. Don’t even think technical or, you know, technology or anything. Just think like process. What. Step one what do we do for step two? What do we do for step three. What step four how many people are breaking at each point and just go on and on and on.

Jason Lazarus00:25:26

Well, and that’s another good segue, because I really wanted to spend some time talking about AI and the changes that are already here and ones that are coming, but as a starting point, the the people that I’ve spoken to have said really the key is, and we have done this internally for our own business, is to identify processes that are administrative in nature, that a machine can do that make you more efficient and profitable, and start to map out how to deploy AI and agents and other technology solutions to solve for that. And so I’m curious from your standpoint, what AI is genuinely good at in personal injury operations right now and what it’s bad for?

Shim Hirsh00:26:27

I’ll tell you what it’s bad at. It’s very bad at talking to clients, and I don’t think it should ever talk to to clients. I think it’s funny most people agree, but they’re so tempted by the, you know, by the by the seeming ROI on on. So I would, I wouldn’t do, I wouldn’t do that. I also think that like the vision of this autonomous AI powered case manager, you know, kind of this agent making decisions and autonomously moving files around and sending things out. And I also think for the pretty near future, that’s also a mistake, really, what it comes down to, if we just take a step back for a minute, I’d say that we’re in a time of what’s called capability overhang, massive capability overhang. The the AI models today are really capable of a lot of things their way more capable than any organization’s ability to adopt this, these capabilities. Right. So the challenge for a business in in getting ROI out of AI, which is my favorite topic, is not is this model capable of doing, you know, task A, B and C? It’s it’s how can I integrate this into my business and operationalize the output? I mean literally the frontier models anthropic OpenAI. They just, you know, in early May both announced what what a surprise at the same day. I wonder how that happened. But they announced a multi-billion dollar investment in integration companies literally deploying their own engineers at large enterprises to integrate AI into the business, understanding that, like, you know, having some intelligent output on a different platform and some other source of truth is not it’s very hard to tie that to in. Right. And so if you just to answer your question specifically, what’s a good at I mean, AI is really good at taking a very specific problem, taking very specific instructions, using very specific tools and producing a very specific output. It’s really good at that. You mean you can fill in the blank on input insert medical document here, look for all these things, extract these things and make it into a medical summary chronology. Like it’s really good at doing these things. And that’s a complicated one. So of course anything less complicated it can handle, it’s it’s the, the operational ingestion of that output. That’s that’s the hard part. I mean, it’s not just a pie problem by the way. It’s not just on, my gosh, only personal injury lawyers don’t know how to do this. It’s it’s industry wide. I mean, it’s MIT just released this this a new study that says that 95% of at least their participant companies could not draw a straight line from their AI spend to any ROI. I mean, the CEO of Uber just said the same thing. Hey, we we burnt our total annual 2026 token budget. We burnt it completely by April. And yet I cannot draw a straight line from that AI spend to any sort of measurable ROI. And that’s why. That’s why I’m obsessed with it. And I spend all my time on the on focusing on not selling somebody an AI technology because it’s just a tool. It’s like, I can tell you a hammer, but, you know, or my favorite analogy, I could tell you an F1 car, but if you don’t know how to drive it, you’re not gonna win the Grand Prix like you’re not. It’s useless to you. And so that’s the part that I think needs strong focus.

Jason Lazarus00:29:58

I think you brought up something, at least to me. That’s key is how does AI integrate with the workforce? How does the workforce use it to either take tasks off their plate that take away from high level legal work, or take away from client interaction, and then turn that into, you know, a better experience, which in turn, you know, ultimately the ROI on that should be pretty clear versus just deploying a tool and hoping that somehow some way that that translates into ROI. I don’t I don’t I don’t see that that is even possible. And that gets back to the idea of what we did, which is mapping out the things that take people a lot of time that a computer or technology or AI can do, and automating and using AI and whatever other means to give people that work each day an advantage and more time to do things at our high level instead of administrative BS.

Shim Hirsh00:31:12

It’s it reminds me of of a bunch of moments at Morgan and Morgan, any sort of technology, it lives or dies by its adoption into a business. Right? Specifically non technology companies, of course technology companies, technology itself drives revenue. But in the industry we work in and most other service businesses, technologies is a tool. And so the adoption like like you were suggesting is is the absolute critical KPI to focus on nothing else. We lived and died by adoption at Morgan. And so I mean, he’s a perfect example of a project I worked on at Better Works. It’s a it was a very small nuance, but massive impact in a difference. Early on we were building this, this system, a bottleneck that that a firm had was not even in retrieving medical records because most most firms use vendors for that, or they have their own network or, and or both. Right. But the the it was the looking through a file to extract the providers, structure them right. Get the address and the name and all the right people compile the packet and merge all whatever docs you need to merge, and then route it to the right place. Hey, do we know this person? Are they in our network or should we send it to a vendor, submit it to the vendor, follow up with our providers and get it right? The whole end to end process was just it was too much admin work. And so we built a system to handle literally the entire process from end to end. And it works fantastic. But the first way we launched it was, okay, well, when you’re ready to have medical requests be made for matter A, create a task to the AI agent and the AI agent will take it on. And the entire we spent all of our time on that end to end system. And it was we did a great job. It does a perfect job, but the entry point relied on a user to to to assign. That’s a behavior change that we’re asking for now. And so what happened? I’ll let you answer the rest of the story. You know what happens, right. So well, you know, once I’m doing it already, I’ll already I forgot, you know. And so what we did instead, as of v2, was we actually built a way to identify all the matters that needed requests and surface them, you know, to. Hey. Yes. Yes. No. No. Yes. No. Edit. Yes. Yes. No. No. Thanks. Right. And so adoption becomes almost irrelevant and we’ll get good enough. We don’t even need that human in the loop confirmation. But that adoption changes any like the technology could be the most sophisticated technology in the world or the most simple. But if you haven’t figured out the unlock into the business, it’s useless, right? So I agree with you on the adoption front.

Jason Lazarus00:33:55

Yeah, I know from our previous discussion that you English between AI that lives outside of the case management software system and the AI that lives inside of it, from talking to enough law firm owners, my understanding is, is that a lot of the AI that is resident in the CMS system today really isn’t great and isn’t firms are not seeing the role from from that. Why why does that distinction matter so much in terms of the the outcome a firm sees from deploying the technology?

Shim Hirsh00:34:39

There’s two things I think, number one, generally speaking, if you’re with with the the way AI is built, I think I don’t claim to be a PhD researcher, but it’s built to consume as much context and reference as possible and draw maps and lines and relationships and essentially produce the the most likely output to make the request to satisfy the request. Meaning by design, it’s producing an average outcome. And so when you’re when you’re when you’re offering a like a platform AI solution, you’re going to get a general response. And the problem in our business and our industry is every single firm has their own flavor, their own playbook, their own rules of the road, their own business model, the very greatly their own, you know, specifics within their jurisdictions. Every county could be different in the same state. And so if you’re not designing an AI output specific to you, you’re you’re then being forced to make changes to it and then it’s already might as well do you just do it, do the raw work yourself at that point. Right. So it doesn’t mean that maybe platforms can’t get there. But the interesting thing with what AI really did for technology building completely shattered the barrier of building technology. It used to be that any company that would start, they would only start if they can, you know, if they had the runway or the vision to build a $100 million AR business because of the amount of money it takes to build technology. And that’s completely shattered. And so everyone has seen their, you know, their nephews and cousins and whatever build these crazy products over the weekend, you know, from their from their whatever from their computer using all these different AI tools to do it. But technology building is, again, not the not the problem. It’s what specific problem I’m trying to solve. How is, you know, is AI even the right answer? Many times it’s not. And when AI generates this output, what am I doing to operationalize that output? Is it just sitting in someone’s inbox? Is it just a beautiful PDF somewhere else or on some some other platform that I’m now begging my people to go and log in and extract and deal with it and move it along, or does it live in my source of truth that then triggers processes that move files and that make people take action? So that’s the that’s the critical part.

Jason Lazarus00:36:59

So right now it seems like there’s three paths for how for deploying the right technology using AI. One is they can build it themselves. And I saw recently Kirkland Ellis I think they’re they’re about to or they may have a building by finding this thing south product or embed AI in their existing tech stack. If you’re consulting with a firm, how how do you guide a firm owner in terms of what’s the right solution.

Shim Hirsh00:37:42

Before figuring out the right way to get to a solution, I like to first figure out whether it is the right solution. So if someone asks me for some for a smart search feature, I usually tell them it’s a dumb ass idea, right? Like what? Unless you can tell me the ROI of that search tool and what it’s going to be used for. And I’m not saying all search is useless. That’s not what I’m saying. I’m just an example of this. I’m thinking of a firm I was talking to. But if you can’t help with Rois, I’m not building that for you. Just doesn’t make any sense. Right? You’re going to. We’re going to spend all the time and all this money and we’re going to build it. And. But let’s say we find a project, right? That we’re right is clear. It’s, you know, on average, this is the most recent project that we that we launched. We’re saving 45 days, days on desk by moving the by making medical record requests a lot faster and quicker and following up more often and frequent and uploading them immediately, receive them back into the back into the system on average 4 to 5 days faster that these now case managers review the files and take action. But as far as once you. So once you find this this ROI, you have the solution for it. The question becomes, well, if it’s if you if you can afford a technology team. Solid, right. And such a technology team and you have the right people, or you can hire the right people, and this is a critical part who can speak your operational language, understand your operations and pain points, and that understands operations in general. It’s a different muscle and can clearly translate that operation, operational need or opportunity into technology building. And that’s the critical part where I say a lot of people struggle. If you can do those two things internally, then you are those people, but they’re expensive and they are defined and you have to convince them to join you and not, you know, for someone else. That’s hard. As far as buying technology, I would say you can buy the the commoditized tools that can handle a very specific, you know, problem with a very specific solution. If you can guarantee that your team will operationalize it and bring it into your system because it’s going to live somewhere else. And put the real trick is, of course, embedding the technology into your existing stack, into your existing software and operations and people. Because again, that’s how you actually achieve any ROI in general. So you got to somehow figure out a way to bring it in-house without having to write build it yourself, which is a distraction. You just want to focus on cases and servicing your clients, really answering the phone when they need you, as opposed to all the other stuff, noise around that.

Jason Lazarus00:40:23

Yeah, I mean, it seems like building it yourself really is in Bible. Maybe for a Morgan and Morgan who’s got, you know, massive resources and, you know, a dedicated department to, to doing it. But I would think for most law firms. You just don’t have that. So if instead you are vetting technologies and you’re getting pitched on technologies, what are the questions that should be asked in evaluating that, that separate something that is real and will actually move the business forward versus something that perhaps is a nice a nicety, but really doesn’t change the way the law firm operates first.

Shim Hirsh00:41:19

And this is like right after you see a wonderful demo, right? It’s just some technologies bloom on the way. How did you get all that information, all the inputs, all the inputs that you needed to actually generate that output? Where did that come from? Did someone have to manually came in? Did you do you have some, you know, some some bot that logs in and extracts much of stuff that’s going to break every two weeks? Like how did you get that information and how does your platform get it. That’s like question number one. Right. The entry point, the number two, can I that that output that you just show me, it’s great for, you know, for a slip and fall trip and fall in New York City. But can I Taylor that that output and the things that I look for to my own case types, my own jurisdictions, my own business model, I would ask, once you have that output, where does it live? Is it on the platform? What does that look like? How do I how do I interact with it? What can I do with it? And then really, you can just ask this last question and answer the other three, which is like, what are your other clients ROI stories and how did you prove it? That would basically tell you what you need to know.

Jason Lazarus00:42:17

So my understanding is, is that you believe the next wave of AI and personal injury space won’t be a product, it’ll be infrastructure inside the tools firms already use, which we’ve been talking about this. I’m just curious about whether you can make that more concrete. What will it look like in a firm 12 months from now?

Shim Hirsh00:42:46

Let’s just say a dream. A dream case, right, a dream scenario. What would it look like? Let me start from the beginning. I think if AI is really deployed the right way, then at a high level, the business, the people and the business focus on the core service of the business, which is supporting clients in their extreme time of need at a very vulnerable moment. And most clients lives when when this injury happens, focus on supporting clients and fighting with the insurance companies and fighting the jesters and defense, defense counsel and whatever it is to ultimately get the best outcome. All the busy work around that should be massively supported by AI. And so start from the top. So an intake agent, I think an intake agent should never have to touch a keyboard, right? I was gung ho about this for years on transcripts and the value of of a transcript and the amount of progress that the technology is made in. The first time I started thinking about transcripts is really incredible, but an agent should never have the touch keyboard. That should just focus on the. Can you imagine having a conversation with someone who’s just extremely emotional and very vulnerable and in their extreme time of need, and you’re running through a script and you’re you’re typing and you’re trying to you’re writing notes. A because your software doesn’t have the fields that you need, you know, like that’s a can you imagine that the experience for getting when you’re talking to that person probably sucks, right? So can you imagine just focusing on and then having all the all that information extracted and put in the right place and analyzed and flagged and the retainer goes out, or the under review case gets into the attorney for, you know, for approval. And just focus on that conversation. Right. You’re focusing on the core service. I mean, you just take that and go down the case, the case lifecycle, a case manager or even the case manager logs into the system. And instead of fielding phone calls of where my case is at and, you know, and whatever provider is asking how much is left in the policy, whether they should treat or not, which is a separate issue entirely. But like, you’re just okay, here’s my list of important things that I need to do on these important cases to move them forward. Right. There’s like a it’s funny, I spent a little bit of time in the construction universe and like my favorite analogy between construction and what we’re talking about here is if you think about it, there is a critical path of tasks that need to get done in sequentially in order to move something forward. And everything else is noise and can be prioritized or deeper. But the critical path of tasks from step one to step two to step three, to step four need to get done and identified. And AI will do a lot of the work around that. And as long as it’s operational, is inside of your right, your your universe, that’s again the critical part that we keep on coming back to.

Jason Lazarus00:45:19

Either maybe easier or maybe difficult to answer. If you ran every personal injury firm in the country and wanted to optimize operations in terms of profitability, and could only change one thing in 2026, what is it? What would it be, and what is the bet you’re making with better works that most people in this space would disagree with?

Shim Hirsh00:45:48

The reason why it’s hard to say to change a one thing is because you have different firms, have different business models and the bottlenecks really, they really, very, very, really greatly from firm firm. I mean, my first four customers of better works had four completely different products, not any zero, almost zero overlap between all four. And not to mention there are four different systems too. So that was a doozy to to to figure out. But if I can roll them all up right into into one case inventory kind of view, I would think about how we kind of talked about this, right. But I would think about, all right, well by high level stage where all my cases, what’s the average age, the average days on desk and in total for each kit for each case in this phase and and average days in that phase. And see what my what my top, top funnel input is. Many sound cases coming in and my bottom funnel top case is going out. And with those three things that you’re looking at, you’ll be able to very easily see where I’m going to get this bottleneck of cases that keep on kind of that get stuck in one place and will continue to expand because I’m pumping cases in the into the machine. But at step four, they’re not really going through fast enough and then dive deep into that area. What I’m betting on on better works, and I’m betting the house is that firm owners and managers, operators will see that AI is not a solution. It’s a tool. It’s an extremely powerful tool. And that the ability to deploy this tool is really part. And it’s like you’re expecting lawyers who are trained to do one thing to now exercise this other muscle. That’s operations and data and technology. And that’s hard. And I’m betting on firm owners understanding that the value when evaluating tech companies or technology solutions is not in the technical ability, it’s the operational knowhow and the business understanding from someone that’s lived and breathed this stuff. That’s where I think the true ROI will start to come out from. And it goes back to that capability overhanging the models can do a lot of things now. They could do a lot, a lot of things right now. And so figuring out or putting together a team of people that can figure out how to inject it in the right places to get the right ROI for me as a firm owner, not for everybody in the industry, for me, what’s my problem to get and have that solved? And that’s where I think the next wave. And it’s not just in pie. I think this is across industries. That’s where again, you see, the frontier models themselves are going after this. That’s where the value is. And you always go where the value is. Good things happen.

Jason Lazarus00:48:23

So final question. It’s very open ended. I always ask it and you can answer it however you want, whether you want to emphasize something you’ve already talked about or bring up something maybe I didn’t ask about. But as a legal technology professional and someone with deep insights and experience, which was evident from our discussion today, what is your view from that perspective?

Shim Hirsh00:48:50

My view is that it’s a really exciting time to be in this industry. Specifically, there’s a ton of attention and a lot of investment on all different sides and angles. There’s an incredible amount of technology that we have. We have access to off the shelf or not. I guess I was always obsessed with creating efficiencies and solving problems and growing and scaling. Anybody can do it. I mean, like any firm can double, triple, quadruple. I mean, with focus and the deployment of these tools, it’s I mean, it’s again, it’s really exciting. What I love about what I do now is that I get to work on so many different problems and different call them firm archetypes, if you will have different characteristics. They have different needs and wants and budgets and spend and different kinds of people in-house. I mean, these are all the different things you have to think about when thinking about a solution. Again, you want to make sure that that the that the tool that that I’m building for these firms that they can again take that car and drive it. And so I guess what I’m saying is it’s just an exciting time to, to be a part of the the massive change in how people work and will will work in the future. And and again, I mean, that’s that’s why I built better works the way I, the way I built it is that I’m not selling a software. I don’t have a login, I don’t have, I don’t sell, I don’t sell AI. That’s not how I position anything that I do. I sell outcomes, so I sit down, I understand the firm and thankfully I could do that pretty quickly. Understand exactly where these bottlenecks are. Sometimes firms, even, they come to me, they know exactly what they want. Sometimes I tell them it’s bullshit like that. Like that search tool, for example. It’s cool to talk about. But that’s not that’s not your problem. Your problem is something else. And so I tie every engagement to the outcome, not to the AI. Right. And so I say that, okay, for every single matter that runs through this particular stage, we’re going to flag it for these 1010 different things. And we’re going to break down the police reports and look for this and look at the medical records for that. And and we’re going to be able to find this. And you’ll be able to let’s just say in one case we’re going to one customer, a perfect example, we’re going to route eight cases to eight teams, be cases to be teams and actually identify be cases that become a cases. Move them up to the A-Team because there’s a big revenue average fee difference between the two. And so tying it to that re means that I’m passionate about the delivery. And that means what’s different about that engagement versus a typical consulting engagement is consulting engagement end on delivery. But the real value starts happening after you deliver. Once you deliver, you realize you have to make you made all these assumptions. You have to start changing things. You have to start looking for different things, or you want to cover more value and you cover. And so owning that outcome, first of all, my customers are much happier with that. They’re much more comfortable because they can’t think about technology and solutions. They just want to know the answers. And so I get to focus on on the technology part. And again, all the time, I’d say 75% of the time that I spend and that our team spends is not on the like, the core intelligence AI feature stuff. It’s on what I call the interpretation layer and the orchestration layer, which is like system design. When does it when do we inject in your system? When do we analyze, when do we where do we take this output? Where do we put it. Who are we emailing. Are we putting it in a dashboard. Are we putting it inside of a matter file? Are we putting it in someone’s inbox? Where’s it going? That’s where the value is. And it’s different for every firm. It’s not a platform that you could just sign up for.

Jason Lazarus00:52:16

Great point. Great way to end today’s episode. If somebody is interested in learning more about better works, has questions about anything you’ve talked about today, what’s the best way to learn and gain contact?

Shim Hirsh00:52:27

I mean, our website IO. You could just email at IO. Text me, you know, I guess my numbers on there maybe. But hey, just reach out to me and we’ll get in touch. Don’t worry. Find me on LinkedIn. Find me anywhere. I’m all over the place.

Jason Lazarus00:52:44

That will include links in the show notes for everybody for today’s episode. And thank you, everyone, for tuning in to this episode of trial or of you. Thank you to for his time. And we’ll see everybody on the next episode.

Shim Hirsh00:52:57

Thank you so much for having me. I love talking shop. It’s my favorite thing. If someone wants to call me just to talk about shop. I don’t care if your customer or not. I love talking, so just give me a call.

Jason Lazarus00:53:05

That’s true. You and I did that before we did this.

Shim Hirsh00:53:09

So my favorite thing. So much fun.

Jason Lazarus00:53:11

If today’s episode gave you a new perspective on how your firm operates or sparked a useful idea, consider sharing it with a colleague and be sure to follow the show so you don’t miss future conversations with leaders across the personal injury. Space trial or view is brought to you by Synergy, a strategic operations partner helping personal injury law firms resolve healthcare liens more efficiently. If you’re looking to accelerate case flow and allow your team to focus on high level legal work that moves cases faster, consider partnering with Synergy. I’m Jason Lazarus and I’ll see you in the next episode.