How AI is reshaping legal services: A conversation with Julien Tanguy, CEO of Lefebvre Group

Welcome to The Heidrick & Struggles Leadership Podcast. Heidrick is the premier global provider of diversified solutions across senior level executive search, leadership assessment and development, team and organizational effectiveness and culture-shaping. Every day we speak with leaders around the world about how they're meeting rising expectations and managing through volatile times, thinking about individual leaders, teams, organizations, and society. Thank you for joining the conversation.

Ahmad Hassan: Hello, this is Ahmad Hassan from Heidrick and Struggles in Paris. On today's podcast, I'm delighted to introduce Julien Tanguy, CEO of Lefebvre Group. Lefebvre Group is a leading European provider of legal and tax services, supporting professionals and organizations with knowledge and decision-making tools. Under Julien's leadership, the group has been accelerating its digital transformation and embedding AI power tools across the organization to better serve clients in legal research and practice, as well as skills development. Just a reminder to our audience, the Lefebvre Group was founded over 200 years ago. It is not just a technological upgrade, rather a strategic and a cultural fit. So Julien, welcome. Thank you for joining us today.

Julien Tanguy: Hi, Ahmad, very happy to be with you today.

Ahmad Hassan: To start things, could you please share with us the group's journey from legal publishing and education towards digital transformation, and how are you launching your AI initiative

Julien Tanguy: Well, Ahmad, as you said, we are a centuries old company. Everything started for Lefebvre decades ago, and today we are still owned by the Lefebvre family, which is our main shareholder, 80% of the capital. So that means this company has a story that has been built on something very important, which is a transmission. And what the family wants to do today, is to manage the transmission of these assets to the next generation. So, how did we build the company we are today? As you said, we are in eight countries in Europe, we are a €600 million revenue company, and for many years, our value was built on the excellence of our content through publishing and training, and content is really where everything started for Lefebvre.

So Dalloz was created before Lefebvre Group in 1845. Lefebvre Group was created at the beginning of the 19th Century. Francis Lefebvre created his first business in training for the tax inspectors in Paris, and at the end of the 1990s, a merger with Éditions Législatives was done to create the leading publishing company in Paris. Then in 2006, Lefebvre and Dalloz merged and then, you know, we went through several revolutions. We started with digital and, you know, years ago digital was CD-ROMs; then we moved to digital portals early in [the 2000s], but then 2023 we [have been] deeply hurt by AI. And we will come back to that, because everything that is happening now in terms of transformation is around AI evolution, and I would say it's more than an evolution, it's really a revolution for the company. Today, paper accounts for 20% of our revenue, so it means that 80% of our revenue is coming from digital.

I think that what has changed for the group is really what happened in 2023 with OpenAI, which was, you know, the first international and generalist LLM. It has changed the way we look at our business, because we have a fantastic database, and you know that generative AI is about how I can use content that already exists, and that will allow the LLM to build new sentences and to answer questions. And because we have this fantastic database, this content that is covering many practices in Europe for law and tax, that we've been able to launch our product whose name is GenIA-L, Generative AI by Lefebvre in other countries. And we will discuss how this revolution, this new technology, is changing not only the product that we sell to our clients, but it has a huge impact on the whole organization of the company, knowing that we were what I call a multi-local group with, you know, countries with a lot of independence.

Ahmad Hassan: Then, how has this changed the content of the discussions at the board level? I mean we're particularly interested in issues like risk ethics, long term value creation, how does that change the dynamic with the board?

Julien Tanguy: Well first, you know, it's about decisions to invest in research and development. As I said, we are very lucky at Lefebvre because we know we have this database which is really the foundation of any AI product. A lot of people are talking about AI today, but the most important thing is really the data. If you don't have data, you don't have AI. So, it has been discussed at group level, at board level, and we had discussions about what we wanted to do in terms of, let's say, a user selling proposition. Because at the end of the day, you don't make a product because you want to please your board, you make a product because you want to please your customers, and because you know that there is potential in the market.

And then what we need to manage at board level today is the time, because, you know, I come from a business where you were thinking in years: When you want to publish a book, you prepare a new book and it takes 12, 18 months. Today with AI, it's not the case anymore. Everything is changing. Every month, every quarter you have new technology that is coming, you have new competitors that are coming on the market, and I can tell you I have many competitors today, much more competitors than I had before. So really, it's all about, you know, the timing. So you need to go fast when you make decisions, and you cannot share everything at board level.

So, we set the rules, we say where we want to go, and then we have regular discussions on a quarterly basis on what happened on the market and what we can do in terms of new products due to technology evolution. And then we monitor, you know, what is happening on the market and what are the expectations of our clients. And I can tell you that the AI product we have on the market today has nothing to do with the AI product we had on the market three years ago. We started with a kind of boosted research engine that was boosted by AI, and we are moving now towards what we call workflows in our business. So, these are use cases that are fitting the business tasks of our clients. So, the name of the game is how can we automate tasks that are managed by our clients, using both our own contents and the content of our clients? Because many of our clients also have their own knowledge base. And the name of the game is to combine their content [with] our content, and to ask AI to work on both to deliver the best outcome.

So at the board level, we have this kind of discussion, and then we are very focused on the execution. Because you can have very good ideas, but now everything comes with the execution and the ability to build the product that you designed before, and this is something we are working on and we try to be educated about what AI allows you to do and what AI will allow you to do in the coming quarters. So [we] have long discussions try to converge in terms of strategy because obviously you have many open questions to answer when you have this kind of topic to manage, and a kind of obsession of what is happening on the market, trying to understand what our competitors are doing, to see if it makes sense for us to add new features to our platform.

Ahmad Hassan: Fantastic. In a few minutes you covered how you're dealing with the board, with the market, with your clients. How about the organization? Because you're leading an organization that has traditionally been based on content creation and you are challenging, you know, there are always the kings in any organization, either you create the content or you take it to the market. So how do you marry the two? Because without the right content, nobody wants to subscribe to your services.

Julien Tanguy: Yes, so it's exactly what you said. First of all, it's very important to understand that content is still something that is very important for me, for my organization, and for my clients. Without those editorial teams commenting on the laws and enriching our database, my AI product has no value. It's still a reality. And even if it's not obvious for people who don't know our business very well, content is still very important. The vision we have for the company is to be the trusted AI-driven legal platform powered by premium content. So yes, we want to be AI-driven. We still want to be trusted, and if we are trusted by our clients today, it's because I have fantastic editorial teams that are, you know, building the database, adding value to our content and allowing AI to run on that. So, my editorial teams are still very important.

Now, if I want to have the best product, I need to combine editorial teams with technology. So I need very strong IT guys—and I have very strong IT guys. And it's not only about the way you use the LLM, it's how you manage your data, how you manage what we call ‘the rag,’ which is the way you digitize the content of your books and of your platforms, and it is [the] combination of those two things that will make a fantastic product. So in terms of organization, it means that people need to work together, which was not the case before. You need to have very close cooperation between the teams, and we changed the way we build our product. So for people who have been in tech for a long period of time, it's of use to work in an agile mode, [but] for us it's something that is totally new. So we changed that last year in order to organize PI planning meetings.

Then we have those meetings every two weeks to track if we delivered what we had to deliver, and if we are able to bring new features to the platform. So this is a big change, not only in the organization but also in the governance of the organization. So it's not only a question of boxes with titles on them, it's really a question about the governance. And then the third pillar that we need to cherish today in our organization is the sales team. Why? Because in my previous business when I was publishing books, I was lucky because I had books that were new every year. It's a kind of millésime.

When you [had] the new Italian tax law or the new Spanish tax law, you know that you will have a new book that will be published and your clients know that they need to buy this book. So my sales team, they were taking orders, they were not selling a value proposition to our clients. When you move to a digital platform like GenIA-L, you need to explain how it works. It's like a ChatGPT, so it's a conversational mode that you need to learn how to work with, and it's really something that takes time. You need to visit your client, you need to make a demo. So it's the best way to understand the tool, but it's something that we need to learn and it is something that I had to explain to the sales teams; you need to change the way you discuss [this] with the clients. And yes, the process lasts a few months, [whereas] before it was lasted a few days. It means you need to visit your client once, then you need to have a second visit. So, in terms of sales, we changed many things in the organization, and we need to be much more KPI-driven than we were before. You need to know how many meetings you need to have with your clients before signing a deal, and the [price] of the deals we sign are much higher today than they were before. You can imagine, you sell books. One book, the value is €100, €200. Today, GenIA-L is more than €200 per month and per user. So if you visit a law firm with 100 lawyers, it means that in terms of proposal, you will send something with a huge amount. And this is something that was not easy for my sales team at the beginning.

Ahmad Hassan: If we go back to organization adoption of change—a lot of the analyses over the past few years say that technological transformation, they often become slow or they fail because of culture more than tools. So how did you approach the cultural aspect of AI adoption, both internally, and also with your customers? Because your customers are used to having paralegals going to libraries and looking at books, now you can tell them that somebody with less experience can do the job equally well as a partner. So how did you handle the cultural aspects, internally and externally?

Julien Tanguy: It's a very complex question, you know, culture management and culture change in the company. I love this sentence from Peter Drucker saying culture eats strategy for breakfast, and I really believe it is so true. So to be clear on that, I am just in the middle of this cultural change for my teams. We started this journey two to three years ago, but the culture has not changed now and we need to accept it. It takes a lot of time, you know, to change the culture of the organization. For me, there are a few things to manage. The first thing is really the communication and the explanation of what we need to change. So, I already said it, but in our business now, time is of the essence, and this is something we still need to work on.

Sometimes we are too slow when it comes to launching new features, when it comes to content management that we need to change. The second thing is that when you want to manage this kind of change, you need to give confidence to your teams. Because, you know, AI is still a big question for many of us:  What will be the consequences of AI on our jobs, on our positions in the organization? Could AI replace me, even as a CEO? I mean it's a question, you can imagine that many people in the organization are asking the same question. And I really believe that the implementation of AI in my company, where AI is not something that is here to optimize processes—I mean, we have a product that is based on AI, so it means that we need to reinvent ourselves almost every month to see what has changed due to technology and, sometimes we really need to adapt and to move fast.

And as I said, this is something that is coming, but it's still in progress in my organization. And as you said, what is quite complex to manage, is the fact that we are in a big change internally and at the same time, the clients we are working with have to manage the same change in their businesses and in their organizations, and nothing is really stable. So we need to find solutions for clients who don't know exactly how they will get organized in one year from now. And everywhere, when we talk with law firms, when we talk with tax advisors, nobody knows exactly what their organization will be tomorrow. And you [had] a very good example: When you were the knowledge manager of a large law firm, you had to manage books, sometimes you had to manage access to digital content, but you had nothing to do [with] the way people are working in the organization. I mean, books were available in the rooms and people who wanted to have access to the books were going to the library and [then] had everything to work with. Now, you have many solutions available on the market, you know that your knowledge management position will be challenged. And I see many organizations where these knowledge management positions are joining the IT teams, because both are very close today.

It is the same as what we have at Lefebvre. I mean content without tech, you cannot do a lot of things with it, so you need to combine and it is the same in the organization of our clients. So we have discussions with them, and it's interesting to see that all the companies don't see AI the same way. In some companies, they see [it as having] less jobs, less juniors, because you can ask AI to do what juniors were doing before. And I have discussions also with law firms that explain to me that yes, they will still have juniors, they will not reduce the number of juniors they have, but the name of the game, how long do I need to turn a junior into a senior? And before, without AI, it took four years, now it could take only 18 months. And why is it very important? It's because the price you charge a junior to a client is not the same as the price you charge a senior to a client. So if you can reduce the time you need to make a junior into a senior, it means that you will be able to make more money with that same person. And you see, so you have two ways of looking at the impact of AI on an organization. At the end of the day, it's too early to say who is right and who is wrong, but you see that you have a negative way to look at it, and also a positive way to look at it.

Ahmad Hassan: Well, maybe they start pricing more for AI augmented lawyers.

Julien Tanguy: Yes, absolutely.

Ahmad Hassan: So Julien, which executive roles have changed the most and which are becoming more critical rather than less critical?

Julien Tanguy: Well for sure, you know, the importance of the role has changed. The executive roles that have changed the most are our product team. So, I have created a new position, a Chief Product and Tech Officer, and this position is obviously now entirely dedicated to developing our AI platform. And it's not only about development, but it's really about product. So, customer feedback and how we need to build the new features on our platform. Then the roles that are more critical today are the people in charge of data because, as I said, it's really where everything starts. And without that, we are nothing. And then, of course, technology. So, what are the new technologies we need to use, and in technology, research and development are key. So, these are the jobs that I would say were not very important in the organization years ago but now are really key.

Then on top of that, people in charge of sales are so important to make our product a success. And obviously, that has changed a lot, and they are more critical than ever, especially in a world where competition is very strong. So, these are the positions that are new and that are important today. And as I said, content was important before, so editorial teams were important before. They are still important. There is no impact on the necessity of having very strong people in content management. So, in a few words, yes, Chief Product and Tech Officer and salespeople.

Ahmad Hassan: And if we just kind of dig deeper in the whole content creation, content management, have they changed the way that they create the content to make it more AI-friendly, or is that not an issue?

Julien Tanguy: It's not an issue, but this is something we have to change. First of all, as I said we have some products with millésime. So sometimes we had products that were updated once a year or once every quarter, and it is a case of something that is very famous at Lefebvre, which is a Memento, which is for tax and accounting rules. So Memento was updated every quarter. With AI, you cannot wait for [even] three months to update the content. You need to update your content on a daily basis. I mean, when you are a lawyer and you ask a question to AI and AI does not know the last law that has been published 24 hours before, you are not updated. And something that was acceptable with paper or with a kind of, you know, digital portal, is not acceptable anymore with AI. So, in a way, we changed the way people are working in content management.

You need to update more frequently, [and] then you need to accept that your first comment will not be as good as comments that you've been working on [over] weeks, but we accept it and then we republish, you know, with more comments, let's say, new thoughts about a law, but we accept it [will] take two to three weeks. But the law, as a raw material, has to be published day one, otherwise your product is not good. So yes, we adjust the way we manage content to be, or to think, AI first.

Ahmad Hassan: Wow. That really reminds me of the transformation of news and newspapers, where you wait for things and the next day for the newspapers, and now you're looking on the internet or CNN to see the video of what happened like 30 seconds ago.

Julien Tanguy: Yes, it's exactly what you said. I mean the parallel between the two businesses is really accurate. When I have discussions with people who are working in newspapers, they've been through the same transformation years ago. And yes, as you said, you have live information everywhere now. You don't wait for the next day or the next week to publish the news, otherwise you're dead.

Ahmad Hassan: So how does it influence how you choose your executive team, how you promote people, how you grow people in the organization? Has that changed over the past year?

Julien Tanguy: No. I mean we try to find new talent because we had some positions that were not in the organization before. But, you know, when it comes to promotion, I really try to have a balanced situation. I cannot, you know, hire 100% of the new positions from outside of the company.

Ahmad Hassan: You lose the culture of the company if you do that.

Julien Tanguy: Absolutely. So it's really a balance in the organization and if I take the example of what we did with the company we have in France, I did something like 50/50. So, as I said, I had no Chief Product Officer in the organization before. Now I have one, and I took someone from another company and someone who has, let's say, two main competencies. So, one is tech, because she's coming from a tech company, but in her career, she has a background within a publisher. So, it was very interesting, you know, to have someone who knows what the publishing business is and who has been able to work in a tech environment. So, this is someone who comes from another company. But then in other cases, for other positions, I decided to keep people that were in place and we are managing the transformation together. And it is really this combination of different talents and different backgrounds that allow me to have a good team today.

Ahmad Hassan: If we have a final question for you, Julien, you know, if you had one or two pieces of advice to give to CEOs going on a similar AI journey, what advice would you give them?

Julien Tanguy: The first advice is—and we said it—I would really work on the impact of this change on the culture of the organization. It's really key. You need to think about how your organization has to change, but not only as I said in terms of boxes, but also to be really focused on culture and ask yourself, ‘what do I need to change’? And the second thing that is key is the governance and the way you're going to make people work together, and for me, that’s really been a big change. You cannot do something within Lefebvre today that has no impact on the other teams. Everything is really dependent on what is happening in other parts of the organization. And then I really believe you need to invest in talent, not only going after talent outside of the organization, but you need to invest in your people, and you need to train them. You need to help them to understand how AI is going to change the way they work in a positive way. Because at the end of the day, as you said, all of us will be augmented by AI, and I really believe that it can be a big, big gamechanger. And then my last advice, if I may, is you also need to be clear with your ambition. I mean you need to change the vision of the organization. You need to see, to understand where you want to go, and you need to be able to set a new vision for your organization. I think it's really important. So it's more in terms of inspiration and in terms of goals, than in terms of pure processing that you need to change your organization.

Ahmad Hassan: Julien, thank you so much for making the time to speak with us today. I really appreciated it.

Julien Tanguy: Thank you, Ahmad, and thank you for your questions.

Thanks for listening to The Heidrick & Struggles Leadership Podcast. To make sure you don't miss the next conversation, please subscribe to our channel on your preferred podcast app, and if you're listening via LinkedIn or YouTube, why not share this with your connections? Until next time.

About the interviewer

Ahmad Hassan (ahassan@heidrick.com) is a partner and member of the Global Technology & Services and Leadership Consulting practices; he is based in the Paris office.