Good afternoon. I'm Mike Walsh, CEO of LexisNexis Legal & Professional. Many of you have seen me in prior seminars, but for those of you who haven't, I joined LexisNexis in 2003 as Head of Strategy and have been CEO of the business since 2011. I'm a lawyer by training, having graduated from Harvard Law School in 1993. After that, I worked at Weil Gotshal, a major U.S. law firm. I also spent time at General Electric as in-house counsel, so I've seen both the law firm and the corporate legal sides of this industry. Today, I will start by providing an overview of our legal business and strategy, including the exciting work we are doing to improve our customer value proposition using AI and generative AI.
Jeff Pfeifer, our Chief Product Officer for North America and the U.K., who many of you will also be familiar with from our prior sessions, will then showcase Lexis+, Lexis+ AI, and Protégé, our next-generation personalized AI assistant, and we will have time for Q&A at the end. Let me start with an overview of our business and strategy. The legal division represents 20% of RELX Group revenue and almost the same percentage of the EBITDA. In the last 12 months to June 2024, our business generated GBP 1.9 billion in revenue, or $2.4 billion. By format, 91% of our revenue is now electronic and continuing to increase by a point or so each year. By geography, two-thirds of our revenue is generated in North America, a fifth is from Europe, and the balance from Asia and other countries.
In total, we do business in more than 160 countries, employ almost 12,000 people worldwide. We have close to 3,000 technologists, several thousand legal domain experts that work on content, product, and operations, and about 3,000 sales and marketing professionals with domain expertise. About 80% of our revenue is subscription-based. Most of our online subscriptions are multi-year, averaging three years in length. We are a leading provider of legal information and analytics to legal and business professionals in law firms, corporate legal departments, as well as government and law schools. We also provide news and business content to a range of customers and continue to provide legal print publications to some of our customers. We are on a strong underlying revenue growth improvement trajectory. When I last spoke to you in 2021, our growth rate had been around 2% for several years.
Since then, we've improved to 3% in 2021, 5% in 2022, 6% in 2023, and 7% in the first nine months of this year. Our underlying adjusted operating profit growth has continued to exceed our underlying revenue growth every year. Our capital expenditure has remained stable at around 9%-10% of revenue. Our EBITDA margin has continued to rise, growing 900 basis points since 2015, reaching 34.6% in 2023. Our objective is to continue on our improved underlying revenue growth trajectory, with underlying operating profit growth exceeding underlying revenue growth. Our strategy has remained consistent. The main driver of our improved underlying revenue growth trajectory is our focus on analytics and decision tools that enhance value for our customers. We are expanding our portfolio of analytics and decision tools that solve key customer problems and provide unique value at key decision points for lawyers.
We are expanding into adjacent use cases across attorney workflows and expanding our solutions globally. We do this by leveraging our comprehensive and accurate content on our global technology platform, which seamlessly integrates extractive and generative AI. This is enabling us to accelerate our organic product development, as well as improve our internal processes, drive cost innovation, and expand margin. The shift to higher-value legal analytics and decision tools is driving our improved revenue growth trajectory. In the chart on the left, which I showed you in 2021, legal analytics, decision tools, and workflow were already a substantial and growing part of our portfolio, growing at high single digits. Since then, Lexis+ and Lexis+ AI have helped us improve our overall growth rate in three ways. First, many of the new AI-powered tools provide another step up in value to the customer, meaning that they are growing faster.
Second, these faster-growing tools are becoming a higher proportion of our revenues, and third, electronic reference is growing faster when delivered as part of our integrated platforms. The development and rollout of new tools like Lexis+, Lexis+ AI, and Protégé, our next-generation AI assistant, will continue to drive the transition toward analytics and decision tools that deliver enhanced value to our customers and will continue to be the primary driver of our growth for another decade or more. We have built these tools by leveraging extractive AI capabilities developed over a decade. Before 2020, our primary solution was our legal electronic reference research platform, Lexis Advance. We also offered a set of high-value standalone workflow analytics and decision tools using extractive AI capabilities that our customers could purchase separately. In 2020, we released Lexis+, an integrated platform that combines electronic reference with these high-value add solutions.
Customers that use Lexis+ are able to access a whole suite of solutions serving many use cases from a single location. This means they can use them more. They want to own more of them, and therefore they spend more with us. As a result, there is a meaningful spend uplift when a customer switches from Lexis Advance to Lexis+. Lexis+ now has users across nine countries, including the U.S., Canada, the U.K., Australia, Singapore, Hong Kong, South Africa, Malaysia, and New Zealand. Last year, we released Lexis+ AI, our integrated generative AI platform that significantly expands the universe of use cases available to our customers. This is driving a meaningful customer spend uplift driven by the higher value add functionality. We now offer Lexis+ AI in the U.S., the U.K., France, Australia, and Canada, with additional countries coming soon. In August of this year, we released Protégé for customer preview.
Protégé is the next-generation generative AI legal assistant that integrates with customer content. It expands the use cases for our customers significantly and enables personalization. As we continue to adopt new technologies and refine our agile processes, the velocity of innovation, product development, and rollout will continue to accelerate. Let me talk in a bit more detail about our different platforms and customer value they deliver. Lexis+ integrates workflow analytics, decision tools, and expands into new use cases beyond our historical core and research. These tools can be used by legal professionals throughout their workday and leverage extractive AI capabilities to drive customer value. Here are some examples. A litigation attorney can use the Legal News Hub to kick off their workday. They can conduct legal research with our comprehensive repositories of case law, statutes, editorial analysis, and more.
They can review clients' briefs and compare them against opposing counsel with Brief Analysis. They can use Litigation Analytics to understand past court and judge behavior to prepare for an upcoming trial, and they can leverage our Practical Guidance content to provide how-to guides on drafting new case documents. Lexis+ AI leverages generative AI to build on our existing analytics solutions and significantly expands the use cases that we serve with our products. Here are some examples of these use cases. Attorneys can save time catching up on legal news by having Lexis+ AI summarize the relevant breaking news and highlighting how it may impact case arguments. Attorneys can understand case law at a deeper level by asking follow-up questions to legal queries in a conversational and iterative manner. Lexis+ AI can help attorneys analyze opposing motions and briefs and identify inconsistencies in their arguments.
Attorneys can upload new litigation data and analyze its impact on a case. It can also automatically draft a brief based on Practical Guidance templates and relevant case law from research conducted throughout the day. Lexis+ AI provides significant value for customers, supporting even more use cases, helping them deliver higher-quality work and saving them time, driving significant demand. Protégé is our next-generation personalized legal AI assistant. It understands an individual lawyer's workflow, tasks, work style, and preferences. It leverages our comprehensive repository of authoritative content and the customer's own proprietary documents, past work, and other firm knowledge. It uses this to provide a personalized experience. Protégé is currently in customer preview with full commercial release planned for early 2025. Our near-term roadmap is focused on integration with customer content and personalized AI for specific legal personas.
Our recent acquisition of Henchman provides us the underlying technology that enables this integration with the customer's internal content. One illustrative example of a use case for Protégé is drafting a contract in a firm style and searching internal documents for applicable precedent. Looking further ahead, our long-term roadmap develops generative AI agents for thousands of specific personas and tasks across various legal domains. For example, within law firms, we'll have agents for real estate attorneys that can draft property contracts, agents for IP attorneys that can compare patent applications, and agents for tax attorneys that can analyze complex estate plans. Protégé builds on the deep integration of analytics and Lexis+ and the significant expansion of use cases leveraging the generative AI capabilities of Lexis+ AI. It represents a step change in personalization for our customers. Comprehensive content and accuracy are critical for creating reliable generative AI-based solutions.
The combination of breadth, depth, type of our content, and value-add entity extraction and linking creates unique and powerful assets that are very difficult to replicate. Our content sets include more than 100 billion documents and records, with 2 million plus new documents added every day from over 50,000 sources. Our citation network with more than 100 different treatments is updated constantly. Some U.S.-based examples of our leading content include federal and state case law, where we have around 25% more coverage than other providers. This content is continuously updated to keep current because a typical case is updated multiple times after its release. We also enrich the data via editorial and metadata enhancements and maintain a highly interconnected citation network with over 150 billion connections. Federal and state statutes include 200 plus years of content and covers beyond what is available through original sources.
Access to statute history and prior versions that are often not available in the public domain is critical for accurate interpretations. Matthew Bender, a trusted proprietary brand of authoritative legal analysis, includes 1,300 plus treatises curated by legal experts that is cited in 153,000 U.S. federal and state cases. We also offer Practical Guidance, which includes tens of thousands of proprietary how-to guides and proven practitioner work drafted by 2,500 leading practitioners across 50 plus jurisdictions and 26 practice areas. Our technology platform is also a key enabler for our accelerating rate of innovation. It is modern and lightweight and allows us to experiment and deploy new solutions rapidly and at low cost.
Our platform has continuously evolved over the years to be relevant, allowing us to adopt cutting-edge machine learning and artificial intelligence capabilities, scalable with multi-cloud environments and deployment, flexible, enabling agile development, and modular with microservices that can be reused across businesses. It sits on top of our leading content and is accessible to our customers through a single point of access. We have been able to seamlessly integrate generative AI into our technology platform. Our approach is allowing us to produce market-leading and continuously improving answer quality, accuracy, completeness, and speed. We have a technology-agnostic multi-model approach that allows us to leverage the best legal domain and open domain AI models for each use case. Our legal retrieval augmented generation, or RAG, platform has been iteratively developed using our in-depth legal expertise.
We began with a semantic search on a vector database, incorporated advanced data relationships via a knowledge graph, and we now leverage agentic RAG that enables solutions for more complex workflows. Our composable, scalable, and multi-cloud infrastructure is seamlessly integrated with our existing global technology platform, accelerating our speed to release new capabilities. Finally, we are prioritizing data privacy and security by ensuring safe use of AI and driving development in line with RELX responsible AI principles. We are continuing to drive cost innovation to manage cost growth below revenue growth, and this is now also supported by generative AI tools. Within our content teams, we have developed the Lexis+ AI content platform, proprietary software that integrates generative AI and automation throughout our editorial operations. This is allowing us to accelerate the publication process, increase our publication output, and maintain our high content standards.
We have also deployed GenAI tools across the software development lifecycle, including a developer coding assistant to generate code, streamline debugging, assist in code refactoring, and more. This is streamlining code production, enhancing code quality, and allowing us to deploy more frequently. To bring this to life, we'll now switch over to our product demo with Jeff, who will show you Lexis+, Lexis+ AI, and Protégé, and walk you through a few examples of how attorneys use our products.
Hello, I'm Jeff Pfeifer, Chief Product Officer for LexisNexis in North America and the U.K., and I'm excited to share a demo of three flagship solutions in our legal product ecosystem. As Mike discussed, our solutions have evolved over time from a focus on reference content to analytics and decision tools, and now to generative AI-powered workflow solutions. Today, we'll walk through the evolution of AI in our LexisNexis solutions.
Starting first with Lexis+, we delivered an integrated workflow solution that combined AI-powered analytics with unprecedented amounts of data and deep insights, which allowed users to work in completely new ways. Lexis+ AI, introduced to customers in May 2023, and it was the natural progression of our platform as it incorporated generative AI, building on a foundation of prior extractive AI investments, and we introduced Protégé in August to customers in a commercial preview program and will introduce it more widely in January. With Protégé, customers experienced a personalized AI assistant like no other, tailored to how an individual lawyer works. In the coming months, we'll also introduce Protégé to additional workflows and to other LexisNexis solutions, expanding exponentially the use cases supported by personalized generative AI capabilities. We'll start today in Lexis+. Lexis+ was launched in 2020 in the U.S. and introduced the first end-to-end workflow solution for lawyers.
As you heard from Mike, this was a major step up for lawyers in productivity as it integrated reference materials and analytics in one platform. You'll notice on the screen the product is arranged by experience, from legal research to Practical Guidance to Litigation Analytics, document analysis, and legal news. Organic development and key acquisitions like Lex Machina in the Litigation Analytics experience and Law360 in the legal news hub are all fully integrated in one workflow solution. Our decade-long investment in extractive AI powered new features and analytic solutions to deliver advanced legal insights for users. In summary, Lexis+ introduced an integrated ecosystem solution enhanced with extractive AI technology and analytics insights, driving significant efficiency benefits and work product quality improvements for lawyers. Next, we'll explore Lexis+ AI, introduced in the U.S. in May 2023.
Lexis+ AI drives new levels of productivity, improved work quality, and economic benefit for legal professionals, utilizing expansive generative AI connected to the largest collection of grounding data from LexisNexis. Lexis+ AI significantly shortens the time spent on legal tasks, allowing the professional to focus on work only they can do for their clients, adding capacity in the day for client interactions, business development, and deeper analysis of legal matters. If you've not seen Lexis+ AI, allow me to orient you to the product. The AI services are organized by task: ask a legal question, draft, summarize a case, and upload to summarize or ask. Importantly, this AI experience is embedded in the familiar Lexis+ experience we just reviewed, allowing the lawyer to easily toggle among AI and traditional approaches to legal work as needed.
Lexis+ AI is unique in that it leverages the deepest collection of trusted proprietary legal content and legal metadata to deliver the highest quality answers and citation references. Let's assume today that I'm an M&A lawyer reviewing dense agreements for key clauses, errors in drafting, or key omissions. Let's take a look at how Lexis+ AI assists in analyzing contracts with a new task that's now possible with generative AI. From the Lexis+ AI landing page, I'll navigate to the upload access point. Here, the lawyer might upload a letter of intent and a merger agreement in order to compare them. You should note that these documents are long, 10 and 46 pages respectively. Lawyers tell us that tools like this in Lexis+ AI are driving significant workflow advantages. Once my documents are uploaded, I will select Ask Questions about my upload.
I'll ask the Lexis+ AI assistant to compare the documents by entering Identify Discrepancies between these documents to ensure that the merger agreement conforms to the deal points in the term sheet. As you can see, the AI assistant clearly lays out important differences between the letter of intent and the merger agreement, including a summary of the key conflicts between the two documents. The business value of generative AI is tangible, and it is positively impacting lawyer work today through improved efficiency and work product quality. Finally, I'm excited to share with you LexisNexis Protégé, an AI assistant that provides a personalized experience grounded in authoritative LexisNexis legal data and, for the first time, the user's own prior work product. As Mike mentioned earlier, Protégé offers a step change in personalization and understanding of the lawyer's workflow.
Protégé further accelerates efficiency and quality of work improvements by anticipating a lawyer's needs and delivering uniquely tailored work product. Allow me to orient you to the Protégé experience, the next generation of our AI assistant. It's designed to introduce an unprecedented level of personalization to Lexis+ AI. On the landing page, for example, users will be welcomed with a customized greeting, but they'll also see access points to ask, draft, and summarize to which they've become accustomed. All services remain grounded in trusted LexisNexis content, delivering the highest quality legal answers. Our new documents task allows users to create their own bespoke databases, store them securely in our system, and execute key AI tasks against the data.
Our expanded profile settings, accessed via the link at the lower left of the page, allows users to review and control their Protégé personalization settings, such as their preferred name, a specific drafting style powered by our multi-model strategy, which applies large language models to specific legal tasks. And we offer other choices, such as preferred legal jurisdiction and areas of legal practice. You'll notice throughout our demo that we've specified California as our preferred legal jurisdiction, and our responses will be highly tailored to California law and California court requirements. Let's explore how Protégé alleviates workflow pain points for the M&A practitioner, exploring further M&A lawyer use cases. Today, we'll assume I've received a marked-up merger agreement from opposing counsel via email with suggested revisions. I can simply drag and drop the document into Protégé for processing.
As an M&A lawyer, I regularly review long and dense agreements, and one of my biggest concerns is finding any potential omissions of key provisions in my contracts. One of the key enhancements Protégé offers is a dynamic set of suggested actions at every step of my workflow, like you see here. These buttons are new and the result of extensive analysis of lawyer workflow. They also bypass the need for prompt creation by the user. Today, I'm checking for missing clauses, so I'll utilize the What's Missing suggested action button to analyze the merger agreement. Protégé instantly generates a report of clauses to consider for inclusion, along with rationale and legal implications of including or omitting each clause.
Protégé correctly noted the absence of a no-shop clause or an exclusivity provision, and the lawyer can learn more about why this provision is important if he or she is unfamiliar with it. I'll ask a follow-up question. Explain why a no-shop clause should be included in a merger agreement where I represent the buyer? Protégé draws on expertly authored LexisNexis Practical Guidance materials, including nine dedicated practice areas that are unique in the market, along with relevant U.S. Securities and Exchange Commission content in order to ensure maximum reliability. In reviewing the response, I see the no-shop clause gives my buyer a stronger negotiating position as it restricts further negotiation with third parties. I can also review further guidance from our proprietary Practical Guidance grounding data used to generate the response. There are documents linked to the full authorities where I can get more information if needed.
Protégé also anticipates my need to edit the document. Now that I'm confident I want to incorporate a no-shop clause, I'll select Edit in Drafting Mode, a new feature to our AI offering. Here, I'll ask Protégé to draft a no-shop clause for a merger agreement. Note that I'm asked if I'd like to draft from LexisNexis content or from my firm's documents. Protégé integrates with my firm's document management system, so it utilizes firm agreements to support research and drafting tasks. I'll select my firm's documents. Our recent acquisition and integration of Henchman, a Belgian legal technology company, allows for Protégé to quickly surface relevant firm precedents and eliminates the need to manually search my firm's repository of agreements.
Prior to acquisition, we assessed a wide range of solution providers and concluded that Henchman's advanced technical capability to indexing customer work product for AI use cases would significantly expand our leading trusted grounding data collection. The no-shop clause we asked for is generated, and we can access the relevant supporting clauses identified from my firm's documents. The firm documents are linked, so I can quickly access the context of the full document in my organization's document management system. I'll next ask Protégé to make the no-shop provision more detailed and buyer-friendly. Protégé combs through thousands of clauses and allows me to refine the language without leaving the browser. As you can see, the clause has been revised, and Protégé is recommending additional language to strengthen buyer protection. I'm pleased with this redraft, and we'll click on Insert in My Document.
From here, I can export to Microsoft Word via our custom Microsoft add-ins for further editing or document preparation. After I'm satisfied with my new draft, I want to update my client on the status of the agreement. I'll ask Protégé to draft an email summarizing the updated provisions in the document. And here, you see a highly contextual email, fully drafted and ready to send after minor modifications. M&A lawyers are often under extreme time constraints to review lengthy and complex transaction documents. As you've seen, Protégé was able to rapidly identify and evaluate key provisions, recommending any missing clauses for increased peace of mind and competitive advantage. Now, with this final client email task completed, Protégé can summarize the provisions for the lawyer's client for further discussion.
Next, we'll navigate back to the homepage and look at Protégé from the perspective of a commercial litigation lawyer representing a corporate client, Globex Technologies, in a contract dispute. In this scenario, the lawyer is in the discovery phase where each party is gathering factual information and conducting legal research to support their case. I know that I want to prepare my discovery documents to send additional requests to opposing counsel. I'll start by uploading a limited set of documents that I've received so far and review how Protégé can assist me with this task. Again, we'll start with the documents task and select Create a New Database for further task action. I'm easily able to drag and drop my file, which begins uploading documents automatically.
I'm able to upload multiple lengthy documents for Protégé to utilize, another example of how Protégé enables significantly faster document review and more accurate document analysis by lawyers. From here, I name my database, Globex Matter, and I could add a description, but today we'll begin with interaction of the database directly. As we saw with a previous merger agreement in the transaction lawyer's workflow, Protégé has many enhancements that tailor the experience to litigation lawyers and their needs. They're also often pressed for time and need help analyzing long, complex documents, as well as drafting different types of legal documents and client communications. Now that my database is created, allow me to draw your attention again to the bottom of the screen where Protégé has provided several suggested actions based on my uploads.
The highlighted options reflect highly relevant tasks in my litigation workflow due to our proprietary method of determining the type of document you've uploaded and customizing your options. More suggested actions that you see on the right shows me an expanded list of AI capabilities to assist with key tasks like research, drafting, and other workflow needs. Each suggestion is personalized to the workflow of the lawyer in line with common tasks, and it's also a progression to agentic AI, the design of complex AI decision systems for legal professionals. With all of the background information available from my documents, Protégé can return outputs that are tailored to the specifics of my case. Since I'm in a legal phase of work called discovery, a first step might be to generate a timeline of events from the uploaded documents in the secure Globex database.
Clicking on the Timeline Action button sends the request to Protégé, initiating a conversation regarding this matter. Events discussed in the uploaded documents are extracted and placed in a graphical timeline view, clearly laying out each item in a chronological sequence. This quickly gives me an overview of events in the matter and valuable insights, such as important procedural events that may trigger additional legal filing deadlines. The timeline view shows that some discovery documents related to an expert witness testifying in the case were recently filed. As a next step, I may want to draft interrogatories or legal questions formally submitted from one party to another. Without leaving this screen, I can simply select Draft Interrogatories as Protégé is aware of my workflow and likely next steps. Protégé will initiate a dialogue prompting me to fill out any additional information needed to complete the draft.
As I represent the defendant in this case, I'll clarify for Protégé that these interrogatories are to be sent to Plaintiff's Counsel. The AI assistant will base its response on both the user-uploaded content and LexisNexis authoritative Practical Guidance form templates written by expert lawyers. This hybrid approach increases accuracy and consistency and reduces the chance of hallucination in our AI-drafted content. Once Protégé completes my draft, the document opens in drafting mode where I can further refine my draft as we saw in the merger agreement example. If I'm happy with this draft, the interrogatories can be exported to Microsoft Word to finalize. However, Protégé suggests further actions to help me in my workflow. After reviewing discovery documents from the case, a likely follow-up task might be to draft a court motion to compel discovery from my opponent.
To get started with this task, I can select another workflow suggestion, Draft a Motion, which will initiate a dialogue regarding the type of motion I want to draft, again as a fully formed document. I enter my response in the prompt bar, A Motion to Compel Discovery. Protégé drafts many common motion types to help save me time. Facts and other data from my uploaded documents can be used to populate required information, so I avoid re-entering or cutting and pasting to complete the motion. Once the motion is generated, I can review the language and the citing legal authority at the end of the response. I'll specifically call out the statute and court opinion references along with our Shepard's Citation integration displayed here as the authority and legal analysis supporting the motion draft.
As Mike mentioned earlier, grounding data coverage advantages and a highly interconnected citation network of more than 150 billion connections drive significant answer quality advantages in our generative AI drafting. Next, I have the option to edit the motion in draft mode, as shown earlier, or I can deliver the motion to Microsoft Word to continue editing. A task that could once take hours can now be completed in minutes with powerful litigation workflow tools built into the personalized Protégé AI assistant. We'll navigate back to the Lexis+ AI Protégé homepage one last time. From the launch of Lexis+ in 2020 to the introduction of generative AI for lawyers with Lexis+ AI in 2023 to the recent release of Protégé, LexisNexis has leveraged next-generation technology to improve the way a legal professional works. Each launch delivered a step change in lawyer productivity and work product improvements.
Our future solutions emphasize personalization of the generative AI experience, driving new levels of improvement to lawyer workflow. Thank you for learning about our LexisNexis AI journey and our innovative third-generation AI solution, Protégé. To summarize today's session, we are improving our revenue growth trajectory by delivering higher-value legal analytics and decision tools to our customers, leveraging generative AI in addition to extractive AI. We are expanding the use cases we address for our customers and providing a seamless, unified, and integrated product experience. We are leveraging our global legal technology platform to accelerate innovation, improve our internal processes, and drive cost innovation. Our objective is to continue on our improved underlying revenue growth trajectory for another decade or more, with underlying adjusted operating profit growth exceeding underlying revenue growth. That concludes our presentation. We'll now switch over to Q&A.
Good afternoon. I'm Mike Walsh, CEO.
We will now begin the question and answer session. To ask a question, you may press star, then one. To withdraw a question, you may press star, then two. At this time, we will pause just momentarily to assemble our roster, and our first question here will come from Adam Berlin with UBS. Please go ahead with your question.
Hi, good afternoon. Thanks very much for the impressive presentation of the products. Maybe I'll ask a couple of questions if that's okay. My first question is, how unique is what you've built here? You've mentioned that there were some other startups you looked at when you had decided to buy Henchman. Where do you think your product is versus some of your competitors? Certainly, it's very impressive to me looking at it, but I don't know what else is out there right now and how advanced other solutions are.
And the second thing I want to ask you is, you mentioned, Mike, in your introduction that people who buy Lexis+ AI see a meaningful uplift in the revenue to RELX. Can you give us any range by what meaningful means? I mean, is that 10%? Is it double? Just help us understand a little bit better what the—I don't think you give an exact number. I know it's going to be different for every customer, but just give us some sense of what that means. It'd be really helpful. Thank you very much.
Yeah. Thank you very much, Adam. Great questions. Let's talk about the uniqueness first. We are super pleased to have an integrated generative AI platform organically built through our offerings. And it's integrated with all the major content sets that we have, including case law, statutes, codes, secondary treatise content, news, Practical Guidance.
It is a technology-agnostic solution, meaning that we can deploy the best LLMs that are available at any given time for any particular use case. We believe this is a leading solution in the market, and we are confident of that. It is a leading solution in the U.S. market. It is leading in U.S. law schools, where we have been out with a fully integrated, comprehensive product in U.S. law schools for three semesters now. It is leading in each of the international markets that we have released to, including five in total now. More are coming on the way. It is particularly unique in the integration of law firm work product. This is a capability that was the number one requested functionality that our customers have asked for.
And the acquisition of Henchman gives us unique technology to be able to access that very broad content set, offering tremendous value for our customers. Let me turn now to your second question on the uplift. You asked about the uplift on Lexis+ AI. Let me bring you back, actually, just take you through the sequence of it. So first of all, we price based on customer value. And the price differs by customer depending on what content sets they're taking, what tools, what situation, what segments, etc. But the first upgrade our customers typically make is to Lexis+ from Lexis Advance. And that's a meaningful spend uplift. And typically, we've seen spend uplifts in the double digits when customers go from Lexis Advance to Lexis+. When they move from Lexis+ to Lexis+ AI, we have typically seen higher uplifts when they move from Lexis+ to Lexis Advance.
And again, it's reflective of the customer value. The generative AI use cases have opened up enormous new value opportunities, and our customers are perceiving that the demand is very strong, and that's reflected in the spend uplift. Thanks, Adam.
Thanks, Mike. Appreciate the help.
And our next question will come from Tom Singlehurst with Citi. Please go ahead.
Thank you very, very much. It's been a great session, so I really appreciate it. And thank you for taking the questions. First one, you may well find this a bit unfair, but I was talking to one lawyer who characterized you as being somewhat behind Westlaw in terms of product development and cross-selling. I was going to say, do you think that's a fair characterization? And specifically against Westlaw, how do you think the Protégé offering differs in terms of proposition? That was the first question.
Then I suppose, secondly, just more broadly, there's clearly going to be a huge value uptick, which is obviously very helpful. I guess the attorneys and the lawyers will be paid back in the form of additional productivity. But if all the main players have all the similar offerings, to what extent do you think over time some of these AI tools will just become table stakes and therefore won't necessarily sustain a sort of growth advantage or growth impact, I should say? Thank you very much.
Yeah. Great questions, Tom. Let's again talk about the uniqueness of our proposition in the market. Specifically, with competitors, we have respect for our competitors. You're best speaking to them about their strategies and their position. We're focused on value for our customers, and the value is significant. We start with the most comprehensive content sets in the market.
We have unique content sets in Practical Guidance, the largest collection of public records, the largest collection of case law. We've got the deepest integration with law firm content in the marketplace. We are the first major company to have released into the U.S. market with the generative AI offering, and I think when you talk to customers or you look at any market reports, they confirm that we've got leading answer quality and leading speed, and when you talk to customers, they will certainly confirm the quality of our offerings. We're focused on innovation, and the speed of our innovation is accelerating. We've got unique agentic RAG capabilities in our underlying platform, and that's enabled us to deliver things for customers that are quite unique in the market that they're clearly reflecting from us in the value.
In terms of long-term sustainability, this is an opportunity that positions us for long-term growth for a decade or more. Very sustainable. We have, as we articulated in the presentation, more than thousands of use cases that we're working on developing as we go deep into the customer workflow. And as you can see from the demo, you need deep domain expertise and excellent content to be able to address those use cases effectively. And it's something that we're uniquely positioned to be able to do with our deep domain expertise, deep content expertise, and the opportunity is significant for a long time.
That's great. Thank you very much.
Thanks, Tom. And again, if you have a question, you may press star, then one, on your telephone keypad to join the queue. Our next question here will come from Nick Dempsey with Barclays. Please go ahead.
Yeah. Hi there.
I've got three more questions, please. So first of all, Mike, at the beginning, you said, "Our objective is to continue on our improved organic growth trajectory." I guess that sounds like your aim is to stay at around seven% organic revenue growth. Of course, that's a really good rate of growth, but could we have some hope that you could continue to improve and accelerate upon that growth rate over time? Second question, if with Protégé you are relying to a degree on the organization and structure of a firm's internal documents, does that add to the risks that there could be teething problems with how useful and accurate Protégé will be from the start? How do you think about those risks? And the third question, when you talk about market-leading onto speed and quality, what are you basing that on?
Is it the Stanford study from earlier this year or some other things that you've done yourselves?
Yeah. Great questions, Nick. I'll take the first question and the third, and then I'll ask Jeff Pfeifer to address your second question at the end about Protégé. Let me first start with our objective on growth. So as we stated, our objective is to continue to improve our growth trajectory. Now, you've seen what we've done from 2021 to 2024. In the past, we were growing at about 2%, and we've stepped that up to 3%, 5%, 6%, 7%. And our objective is to continue to improve that. We have a lot of tailwind behind us with our launches of Lexis+, Lexis+ AI, and Protégé, and the trajectory is strong.
Now, given that our business, as you know, is 80% subscription, largely with three-year contract terms, the results of that growth and that momentum tend to come through gradually. But our objective is always to continue to improve and accelerate that growth. Let me address your third question about market-leading answer quality and speed. So first of all, we do significant testing on this. We've got a lot of metrics that we look at. We look at speed on every different dimension: speed of answer, speed of result. And we test both ourselves and through third parties. We test others in the market. And we do the same thing with answer quality. And we have panels of attorneys who actually evaluate and give us scores on the answers. And we look at these metrics daily, weekly, and we're constantly driving and improving them.
There are many third-party studies in the market, too. Some of them are public. Some of them are not. I've gone through all the studies, and they basically confirm that we have market-leading answer quality and speed. And it's a critical thing for our customers because at the end of the day, you need a high-quality, accurate result, and you want it quickly. Let me now pass it over to Jeff Pfeifer to address your question about Protégé.
Thank you, Mike. And thank you for the question. Related to internal data collections, I alluded in our presentation to the technology advantages brought with the Henchman acquisition. The approach that Henchman uses was very aligned with our own in that it works to organize internal firm data collections, adding enrichment value.
That could be things like structuring data, adding metadata, and filtering highest quality data for use as grounding data in an AI answer. So what intrigued us so much was the ability to take a very large document volume from a client and put it through a technology-driven enrichment process that adds value in that processing and ultimately improves the quality of the data availability when we generate an AI answer. So customers tell us, of course, that their data and their precedent information is critical to them for AI use cases and that enrichment process, we believe, allows us to select the very best data and the most high-quality data as we are leveraging internal firm work product as grounding data for a generative AI answer.
Thank you, guys.
And our next question will come from Natasha Gupta with Cantillon. Please go ahead.
Hi.
The question we had was, how should we think about why electronic reference is growing faster on integrated platforms? Is electronic reference something that most people or most legal firms should be having, regardless of whether or not they have Lexis+ or Lexis+ AI, or is there an improving advantage to having those electronic databases as well over time?
Yeah. Great question. So electronic reference has increasingly been integrated into the workflow of our attorneys, as illustrated in the Lexis+ use case. So historically, research was something that was a very specific task, and the information and content that came with it was very specific to a research assignment. Over time, the content that has supported research has now been integrated into many other use cases. For example, when an attorney comes in and starts their workday, they start it in our Law360 news service.
Our Law360 news service. You pull up a news article, and that will link you to a case or statute information that you're getting. So you're going to be using more of it. We have tools that will analyze your briefs for you. And when we analyze your briefs, we'll cite you to different cases and suggested information that's research information that you might want to use in an analytical tool. And so the interlinkages that we have throughout the business into our workflow solutions in Lexis+ and now Lexis+ AI is actually pulling stronger demand for that reference content. So as that reference content gets integrated into our workflow solutions, it has higher use, higher usage. Our customers use it more, and therefore they want to spend more with it. And so the category starts to grow faster as it gets integrated into our other solutions.
And that will conclude our question and answer session. I'd like to turn the conference back over to Mike Walsh for any closing remarks.
All right. Well, thank you, everybody, for attending. It was great for you to be with us. Any follow-up questions, please, you can contact Investor Relations at RELX. So thank you again, and goodbye.