Michele D. DeStefano, a professor of law at the University of Miami, spoke with Mark Smolik, the chief legal officer at DHL Supply Chain Americas, about how to add value in the age of AI.
Michele D. DeStefano: Mark, you’ve always taken an innovative approach to data and knowledge management. Can you talk about that approach—even before AI—and how you knew, so far ahead of the curve, that knowledge management was going to be king?
Mark Smolik: When I joined DHL’s legal department in 2009, we were a team of five. Today we are 91, spread across multiple countries. Back then, whenever I asked, “What’s the status of a particular matter?” or “Where is a file?” I was always pointed to outside counsel, and it really surprised me. We were too dependent on outside counsel. We had a handful of files internally, but nothing was organized in any meaningful way. I knew we had to change how we operated. I met with my team and said we were implementing a filing system and I expected our information to be readily available in-house.
I knew that if we were going to make a real impact on the organization, particularly when handling commodity-type matters like procurement contracts, leases, and routine employment and related claims, we had to get organized. We literally started with paper, some filing cabinets, and file folders. We organized our information by corporate matters, M&A, employment, and so on. It was basic blocking and tackling. We never strayed from organizing the information that flowed through the department, which I knew was critical, especially as we grew our team over the years. A couple of years later we put in our first document management system, and our digitalization journey was underway.
We had the foresight to say we needed information at our fingertips, readily available.
Mark Smolik, chief legal officer, DHL Supply Chain Americas
Shortly thereafter I vividly recall asking my associate general counsel, the head of employment, to pull every employment claim over a five-year period and put together a report on what was driving our claims, what are the costs we are incurring, and where the claims were emanating from within our organization. It was largely a paper journey, and a laborious one, but the results ended up in an Excel spreadsheet that gave me real insight I could share with the business. For the first time I could report to them on what a claim was costing us on average, where in our operations the claims were arising, and how many claims originated from what part of our business. The data helped us refine our approach to proactive risk mitigation. It was the first time I had meaningful data that drove insight that led to changes in how we were handling those claims and training our business.
There was also an element of dumb luck in my quest for organization. We had the foresight to say we needed information at our fingertips, readily available. But, we had no idea that how we were going about organizing our data would lead to us being well positioned to leverage what we all have come to know today about the features and benefits of AI. In today’s vernacular, we were structuring and tagging our data years ago, but we had no idea how well that would position us to leverage AI.
DeStefano: You’ve made an open declaration that your department is going to be 50 percent more efficient in two years than it is today. What does that look like? And where is AI really adding the value?
Smolik: We’re already realizing many of those efficiencies. For me, AI is about augmenting the productivity of the team. Our business continues to grow quite significantly. But that growth does not mean our team will grow at the same pace. We have to continuously improve, finding new and creative ways to deliver services faster, better, and cheaper. Any organization seeking to extol the benefits of AI needs to first get their data organized. We are fortunate to have organized data that serves as the critical foundation of the AI initiatives we have underway.
Four years ago, we built an in-house AI employment (EEOC) claims tool for a modest investment (less than $150,000). Before that, we were sending EEOC matters to outside counsel at a cost of roughly $37,000 per claim. Now it costs a fraction of that, with AI agents drafting the initial responses and team members reviewing them while keeping much of the work in-house.
I foresee that pragmatic department leaders will conclude: ‘I need someone to not only organize and manage my data, but to tell me what that data says.’
Mark Smolik
The past couple of years have been busy with M&A, an area ripe for the use of AI. In a number of our deals, we conducted due diligence leveraging Eudia’s Insights platform while having our external law firm prepare their traditional due diligence report. On one of our most recent deals, our outside counsel informed me, “I hate to admit it, but your AI report is good; no changes.”
We are also driving efficiencies through our department’s intake tool, which we’ve had in place for the better part of 11 years. It identifies the type of matter and routes it to the right person based on capacity and expertise. I’ve shared with my CEO my goal that when the business sends to legal a request for assistance or a document through that portal—an NDA, for example—before the assigned attorney or paralegal even opens the new assignment, an AI agent has already reviewed the document, marked it up against our knowledge bank, and then routed it to the attorney or other professional in the department to whom the matter was assigned. A review that once took an hour now takes minutes.
We made a great investment three years ago in our first data scientist, and we’ve added more since. That, to me, is one of the many shifts in the composition of roles in a law department that will take place in the months and years ahead. I foresee that pragmatic department leaders will conclude: “I need someone to not only organize and manage my data, but to tell me what that data says. That is where a team member with data experience will be of significant value.”
DeStefano: Do you think law firms should have a data scientist, too?
Smolik: Yes. The firms that take a genuinely proactive, business-minded approach—the ones that say, “I can’t leverage effectively the features and benefits of AI until I have my data in order”—are the ones that will be most competitive in a hypercompetitive market. In some ways it matters more for the firms than for the departments, because law departments don’t compete against one another; law firms do, and they need to find their competitive edge. The effective use of AI laid on top of organized and structured data will surely be one of those competitive advantages that will help drive measurable and sustainable value for both buyers and sellers of legal services.
My team knows exactly how much revenue our business needs to generate in order for us to have a dollar to spend.
Mark Smolik
DeStefano: How do you define that value? How do you know when you’re delivering it?
Smolik: It’s not about how many cases you win or lose, or how many contracts you close in a month. About eight years ago, my CFO said to me, “Traditional law department metrics don’t mean much to the business.” I’ve heard that from others over time. I don’t report on them. Instead, I manage to results and to the perception of value held by the person I am supporting. Their perception of me and my team is their reality. We are the ones who create that perception through the service we provide, the commitment we make in supporting their business, and the desire to help them as best we can. One of our retired CFOs advised me: “The people who make decisions about your value are the people you support, which at the board level are me, our CEO, and our business leaders. If any one of us walk out of a meeting with you and say to each other, ‘He gets it—he thinks like the business, he understands what we are seeking to achieve and will find a way to help us achieve our goal ethically and with an eye on the P&L’—then that, Mark, is creating value.” I’ve never forgotten. It has kept me focused on the value proposition.
DeStefano: What he was really telling you is that they value you as a business counselor and partner, not just someone minding the legal department’s costs. Still, aren’t they asking you to cut? Every corporate legal department seems to be under that pressure now, especially with AI.
Smolik: No, they’re not. My team knows exactly how much revenue our business needs to generate in order for us to have a dollar to spend. I stress we are businesspeople with law degrees and legal training. The team is vigilant about fiscal discipline, keenly focused on service excellence and the protection of our P&L. Everyone in my department understands that financial reality. When legal starts talking like businesspeople who happen to have law degrees, the business really appreciates it.
In 2009, we sent about 85 percent of our work outside. Today, roughly 80 percent is done internally.
Mark Smolik
We used to run a laborious, intense budget review with our CFO. I haven’t had a review last more than 15 or 20 minutes in years, because we’ve shared our business model and approach with our CFO and our CEO. We lay out our historical external fees and where they’re heading over the next five based on historical trends. We then share our team financial data, which includes what it costs to employ each person on the team, fully loaded, versus their counterpart at a firm and what it costs to run the department. Our business has challenged me to the extent we can keep work in-house. That means investing in an internal team that continues to rely less and less on external counsel.
In 2009, we sent about 85 percent of our work outside. Today, roughly 80 percent is done internally.
DeStefano: Let’s turn to the human side. What is essential about human lawyers, and how can they add value in the age of AI?
Smolik: Listen—seriously. As lawyers, we are trained to leverage our knowledge to help people in need. I’ve experienced many lawyers who try to impress people with the level of their knowledge, which was oftentimes more about the theory of the law than its practical application. I can’t remember an instance where someone in the business cared much about some lawyer who tried to convince them they were the smartest person in the room. People don’t typically care. They come to us with a problem, and they look to us much the way they look to their doctor: to listen and then leverage their experience and knowledge to diagnose and help them solve a problem.
Data can help guide you in that decision process (so, too, does your experience level), but you still need the human element, the judgment, to ultimately decide what to do.
Mark Smolik
Our business operators know there’s a risk in every action they take. Our job as members of the legal team is to help them figure out what is the real risk. I can’t tell you how many times a businessperson has said to me, “I’m told there’s a risk if I take this action or that action.” And I say, “Of course there is. There is a risk in everything we do. Let’s talk through what you want to achieve, what is driving your decision (or the problem), and what processes or disciplines you already in place that are relevant to your decision.” I am constantly evaluating whether the risks presented will put the P&L in jeopardy or if it is more of a theoretical risk. Of course, there is more to it than that, but more often than not, by the time you’re done evaluating the risk, you realize it’s manageable and oftentimes mitigated (at least in our organization) by the sound operating disciplines or processes already in place.
That’s where the art comes in. Out of law school, you’re trained to flag risks, and it takes time to learn which risks genuinely threaten an organization and which don’t. This is where data and AI can help. Imagine being asked for your advice about a dispute or claim and you turn to the department’s database and query it to show you every claim that fits what is being described to you. The search comes back with a summary that reveals there’s been one minor claim. That type of insight can give people a level of comfort that there’s objective data behind the decision regarding the risk itself. Data can help guide you in that decision process (so, too, does your experience level), but you still need the human element, the judgment, to ultimately decide what to do.
DeStefano: Are you getting that kind of service or counseling, including using AI, from your outside law firms?
Smolik: We’re not, although one of our firms (Ice Miller) is an exception. I’ve heard many of our panel firms extol the virtues of AI and cannot help but think: We spend a great deal of money with you, and I’ve never heard a word about how you are using data and AI to deliver to us faster, better, and less expensive services. Fees continue to go up year after year. I have seen no cost efficiencies from our firms. I certainly don’t see cost efficiencies in their invoices, and I don’t see cost reductions or less time billed on our various matters.
To drive some of these efficiencies, we’re pushing our firms to change.
Mark Smolik
The only conclusion I can draw is they are using AI to drive efficiencies that seek to preserve, protect, and grow their business and margins. I’m all for firms making money. But don’t go out and tout to the industry what you are doing with data and AI while leaving your clients wondering what you’re actually doing with AI, data, and technology to deliver services better, faster, and cheaper. Our invoices aren’t going down (or at least flattening), while your rates increase every year.
DeStefano: Firms will say they haven’t seen the benefits yet because so much human review remains. Why aren’t they being more proactive and having that conversation with you?
Smolik: I’d welcome that conversation. But they are not taking place. So, we are taking action on our own.
To drive some of these efficiencies, we’re pushing our firms to change. We started with our two largest firms in North America. About four years ago we developed a concept, now live, that we call our Collab Hub. The vision was simple: any data the firm has on our matters is ours. It’s DHL’s data, and it’s all tangled up in their environment. I want it segregated. They can keep a copy in their own ecosystem, a data lake, but I want it in a DHL data lake that the firm is required to access when new matters are assigned to them by our team. The data in the “DHL lake” has to be leveraged to evaluate if a matter is similar enough that previous work can be leveraged and reused. My position is, we’ve paid for the work already. Why should we pay for it from dollar one billed every time a new and similar matter is assigned? Recycle and refresh work instead of billing us as if you’ve never handled similar matters or claims for us.
We included in our billing guidelines a requirement that when a new matter comes in, the firm has to check the hub for similar work. If it exists, they are required to leverage it out and update it, but they cannot bill us from dollar one, and we’ll know if they do.
When it comes to a department’s digitalization journey, the first step is to get its house in order.
Mark Smolik
DeStefano: A lot of legal departments aren’t where you are, and the firms are often even further behind. Where does the data actually sit, and what’s your recommendation for getting control of it?
Smolik: Corporate law departments first must settle one question: Who owns the data their firms hold—the firm or the client? The firm may control it, but it’s your data. Go to your firms and have the conversation, starting with the premise: “It’s mine. I want it segregated, and I want it organized, structured, and tagged so I can leverage work performed.” The firms already keep it in a controlled environment. They know where the data is (at least they should).
When it comes to a department’s digitalization journey, the first step is to get its house in order. Develop a business strategy (some call it a targeted operating model); then aggregate, organize, and structure your data. This may seem daunting, but it’s not once you have a strategy and a plan as to how you want to use your data and for what. Don’t be tempted to first go out and buy the shiny new AI tool. You won’t be able to use it effectively unless you have the data to support it and use it as designed.
DeStefano: One question everyone’s been thinking about: these tools help senior lawyers who have the judgment to work with them. But what about junior lawyers and students still in law school? What’s your advice to them?
Smolik: In our industry, we get paid for our judgment, which is driven by our knowledge. In many ways, knowledge is power. The more of it we accumulate, the more influential we can be in delivering legal services.
People like to deal with people who listen.
Mark Smolik
Many people are speculating that AI will eliminate jobs. It will in some industries, but I think it will create more jobs than it eliminates. I am encouraging people to embrace it and use it to augment their productivity. I recall many years ago, when I was a summer associate at a law firm in a class of eight very bright and eager associates, I surely wasn’t the brightest, but I became the one who knew how to use LexisNexis better than most. Others seemed to resent it when more and more members of the firm turned to me for support on their research projects. I realized then, and have never forgotten, that the more expertise you build in using available tools to drive efficiencies, the more valuable you become and the more people will rely on you for your expertise. So, when it comes to AI, learn how it “thinks,” how it operates, and how it can be optimized to help you as a legal professional.
Second, people like to deal with people who listen and don’t carry the attitude of “I’m the lawyer; you’re not.” If that is you, then drop the attitude. Humility is critical: you’re no different from anyone else, and you’re there to help someone solve a problem.
The greatest hindrance to advancement isn’t a lack of talent. Oftentimes it’s founded in attitude and a failure to keep learning. You never want to hear people say, “Brilliant lawyer, but I don’t like dealing with him or her.” Don’t be afraid to take criticism, even when it’s negative. It’ll teach you more about yourself than you may already know.
DeStefano: To wrap up, what’s your next step?
Smolik: I’ve announced to DHL I am retiring from the organization and will be leaving toward the end of this year. It’s been a remarkable 17 years. I’ve had the privilege of working alongside incredible people with deep values and strong integrity. My CEOs and CFOs over the years extended to me the autonomy to do some remarkable things, including investing in legal operations 16 years ago, when many of my peers looked at me sideways and said I was crazy to do so. I’ve been very fortunate. But I’ve got a bit of an itch to do something broader. I’m not moving on from the industry. I still have a number of things I want to do to help drive meaningful change in an industry that has been so good to me. More to come on that soon.
Mark Smolik is chief legal officer at DHL Supply Chain Americas.
Michele DeStefano is a professor of law and the Larry Hoffman Greenberg Traurig Business of Law Chair at the University of Miami; affiliated faculty at Harvard Law School Executive Education and IE Law School; and the founder and director of LawWithoutWalls.
