Start With Why: How Legal Professionals Are Embracing Design Thinking
Author: LDI Team
The Legal Data Intelligence model offers a framework for applying technology—particularly AI—in a more strategic and defensible way to address the data challenges that shape legal work. Yet for many legal teams, the instinct to adopt AI is often accompanied by a rush toward tools and systems. In that haste, a more fundamental step is frequently overlooked: defining the problem itself.
At Legalweek 2026, LDI founding member Scott Milner, partner at Morgan Lewis, moderated a discussion titled “Design Thinking in Legal: A Modern Approach to Tech Adoption and Problem Solving.” The panel included Catherine Alman MacDonagh, CEO and founder of the Legal Lean Sigma Institute; Jessica Escalera, head of legal operations, Americas, at HSBC; James Vinson, senior director of eData at Morgan Lewis; and LDI Architect Stacy Lettie, chief of staff to the general counsel at Organon.
The discussion covered instructive case studies on applying design thinking principles in a legal context, the flawed assumptions legal teams might make when integrating AI into their processes, the importance of taking a ground-up approach when selecting technology capabilities, and how to unlock the full range of possibilities AI presents in rethinking legal work.
Below are some key takeaways from the session.
What Really Is Design Thinking?
“What’s fundamentally different about design thinking is that you start with the people first. You don’t start with the problem; you start with the problems that people are actually experiencing. Not what we think the problem is, not what our biases have filtered for us. It’s the transition from knowing what we think the problems are to learning the true needs of stakeholders,” MacDonagh said.
A key attribute of design thinking is bringing every stakeholder onto the same page about the problem they are trying to solve. As work demands pile up, deadlines loom, and workplace silos create communication and information gaps, legal professionals often find themselves enmeshed in projects without the opportunity to critically examine why the project is being undertaken in the first place.
Design thinking is an antidote to that default condition. “It seeks to articulate with specificity the ‘why’ behind the project and describe in detail what success looks like,” MacDonagh said.
People, Process, and AI Literacy
That laser focus on truly understanding the problem is what separates successful technology implementation projects from unsuccessful ones.
Escalera elaborated on two principles that shape her view of AI adoption. The first is taking a human-centered approach. “You have to start with what you are bringing to your people. What is the story you are trying to tell? For us, the AI story isn’t about using a specific tool. It’s about moving a legal team up the AI literacy curve. It’s about understanding people and how to reach them.”
The second principle is understanding the process as it exists today. “People shy way from looking at their processes. But it really is the most important part. If the process is broken, putting AI over it is not going to fix it.”
Lettie echoed Escalera’s emphasis on understanding the nuts and bolts of the process as the essential starting point in any technology transformation project. “Years ago, in my first legal operations role, I was tasked with implementing a legal intake system. I went ahead and designed it with the technology provider in a way that I thought would solve everyone’s problem.”
However, Lettie began noticing issues with the system during early user testing. “I realized that I was trying to fix problems that didn’t exist. I had recreated the process and made it even more complicated.”
Regardless of the use case, the right approach, according to Lettie, always begins with taking stock of the process. “You have to do the legwork first so you can identify where the actual problems lie. It’s your responsibility to detect and analyze the problem. You can’t rely on a technology vendor to identify it for you.”
Building Bridges in the Disputes Space with Legal Data Intelligence
Detailed case studies and KPIs for LDI practitioners on closing gaps and fostering cross-functional collaboration.
Read the ArticleThe ROI of Diagnosing the Problem and Unearthing its Root Cause
MacDonagh cited a case study from her recent book, Lean Six Sigma for Law, Second Edition, in which a global law department at a major company was dealing with a billing problem involving outside counsel.
“The bills were getting rejected, and none of the lawyers from either side knew why.”
Having the right people in the room helped uncover the root cause. “We gathered everyone, including the billing and finance teams from both sides. Ultimately, it was the people closest to the problem—the finance team—who could tell us why the bills were getting rejected. It was because they had the wrong timekeeper setup and the wrong matter ID.”
That “aha” moment enabled the team to reduce bill rejections by 40% overall—72% from fixing the timekeeper setup and 56% from correcting the matter ID. “This was after one day of people working together, going through the design thinking process with in-person facilitation.”
Designing Brainstorming Sessions That Translate into Outcomes
At a time when legal teams are expected to leverage AI and demonstrate fluency, they may feel pressure to embark on transformation efforts without first understanding how AI can add value in concrete terms.
Lettie returned to first principles. She wanted her team to understand how AI works and to equip them with practical insight into how it can be applied in legal workflows.
To that end, she organized a four-hour workshop over two days. “ I had our IT department explain how AI works. Then we had a session where Scott [Milner] and his team discussed how AI applies generally within a legal department and how it might be useful to the practice areas in our particular legal department.”
Following the sessions, Lettie led a brainstorming exercise in which team members were asked to create a “wish list” of how AI could support their daily work.
“We asked them not to worry about whether we could build it or whether we had the budget for it.”
The team generated over 20 ideas. “It’s only been a year since that exercise, and we’ve already made progress on 4 or 5 of them with several more in the pipeline.”
The Importance of Prototyping
“We’ve talked a lot about ideation and brainstorming, but not as much about prototyping and experimentation,” Vinson said. He cited a software contract renewal where he moved quickly without getting sufficient input from stakeholders. Reflecting on the experience, he identified the lack of experimentation as a critical gap, noting that even a simple prototype could have revealed flaws in his assumptions much earlier.
“Had I done that, I would have very quickly gotten feedback on where my thinking was off the mark on this renewal and how people were actually using the tool in ways I didn’t realize.”
AI as an Opportunity to Rethink Processes
As the legal industry moves to integrate AI into its workflows, it may be overlooking a broader opportunity: to rethink those workflows altogether.
“I think it’s really important to go back to first principles and design thinking,” Escalera said. “Too many people are thinking about how they can layer AI over their current process. Not enough are asking, ‘What if we didn’t do things the same way we’ve always done it?’”
That distinction—between layering technology onto existing processes and reexamining those processes from the ground up—is where design thinking becomes most relevant.
“I would encourage practitioners to step back and apply design thinking to workflows that have existed for decades, including discovery and contracts,” Escalera said. “If we start thinking from first principles, there’s a very good chance we will interact with AI in fundamentally different ways than we do with current technologies.”
(Editor’s Note: Legal Data Intelligence was a dedicated track at Legalweek 2026. Check out recaps of the other panel discussions under the LDI track below: