Right, good morning, everybody. Thank you for coming. For those who don't know me, my name is Julie Southern. I'm Chairman of RWS. I think it's a very overused phrase, but nevertheless a very true, true phrase to say that change has never been happening faster than it's happening now. It's also true, I think, to say that with as much change as we're happening now, it's very hard for us all to imagine kind of the impact that that's going to have on us as individuals, on society, on how businesses work. Out of disruption, usually good comes and society progresses, and businesses find new ways of doing things, new approaches to the market, and everything moves forward. Now, I say that, and I don't mean that everything moves forward in a simple and seamless journey. It's hard.
It often faces setbacks, and it probably takes more time than many people think. When I look at RWS and I look at the work that Ben and Candy and the team have been doing, I think the thing that we are really excited about is that most of the macro trends that we see at the moment, we think play into areas that we are already good at and spaces that we have every right to succeed in. We're not going to succeed in them today without changing, but we have a clear view, which you will hear from Ben, on how we're going to change and where we need to change.
I'm really delighted to be able to hand you over to Candy and Ben, who are going to talk about both the progress we've made in the first half of this year and then, excitingly, the journey that we're planning to go on over the next, certainly over the next six months when you'll get more detail, but, over the next year or so. Enjoy it. I hope you find it really informative. I'm sure you will. Thank you for coming.
Thank you, Julie. Thank you and welcome everyone. Good morning. Very excited to be here with Candy. Today we will demonstrate how a new technology-first strategy supported by a leaner, more agile organization will accelerate growth and improve profitability. Our focus on embedding AI and automation at the heart of our operations, coupled with strengthened execution across all business units, will generate higher quality, more predictable earnings, and deliver higher value for our shareholder, long-lasting value for our shareholder. We're going to start with Candy, and I'll come back to detail the strategy. Thank you.
Thank you, Ben. Let me take you through the financial performance for the first half of 2025. We reported revenue of GBP 344 million, reflecting organic constant currency growth of 1.4%, marking our fifth consecutive quarter of growth. Gross margin declined 240 basis points to 43.3%, impacted by business mix and investment in TrainAI. We expect improvement in the second half due to revenue growth and our ongoing cost reduction programs targeting approximately 44% for the full year. Adjusted EBITDA was GBP 38 million, delivering 11% as a percentage of revenue, and adjusted profit before tax was GBP 18 million, slightly ahead of our trading update. This translates to an adjusted basic EPS of 3.6 pence. Capital expenditure was 3.4% of revenue, down from 6.4% in the full year 2024. Even without the change in capitalization treatment in the first half, CapEx would have reflected a decrease as previously guided.
Finally, despite the lower profit in the first half, I'm pleased to confirm that the board has approved an interim dividend of GBP 2.45, consistent with the prior year. Moving to the income statement now and providing some more color on the key items. As I said, gross margin declined by 240 basis points. Whilst ongoing efficiency efforts are gaining traction and have contained inflation and the price pressure felt in our core localization businesses, we have experienced some adverse volume and mix effect across the business. There are several drivers here, so let me take a moment to unpack it a bit. Firstly, we've experienced weaker trading in our regulated industry segment, which has not only resulted in a lower margin within that segment, but has also driven an adverse mix impact for the group.
Secondly, we've seen accelerated growth in our TrainAI business, which has had a dilutive impact on the group margin. We're currently investing in capabilities to support the growth and ramp up, and would expect to see margin expansion in this area over time. In addition, as we explained at the time of the trading statement, we've experienced some one-off efficiency shortfalls, particularly with two clients within language services, with one client as they ramped up volume on our Evolve MT engine and another client who's also ramping up fast, but is transitioning to a newer toolset of their own. We expect to see resolution of these issues in H2. Finally, the divestment of PatBase has had a small impact on the margin, as has FX, as the pound continued to strengthen against the dollar and euro.
Admin expenses before adjusting items of GBP 128 million increased GBP 16 million year- on- year, reflecting the increased amortization linked to historical investments, as well as an updated approach to the capitalization of software development, with a larger proportion of our investments being expensed in the year. Inflationary cost increases and investments in growth and core capabilities were offset by ongoing cost efficiency programs and cost control initiatives. Our headcount in overheads reduced 5% from September to March, and we continue to focus on this as we drive simplified processes and greater automation. Adjusting items of GBP 31 million are in line with the guidance given in December.
These include amortization of acquired intangibles of GBP 22 million relating to the past acquisitions of SDL and Moravia, exceptional items considered one-off in nature, encompassing transformation and restructuring costs of GBP 5 million, acquisition-related costs of GBP 3 million, including the Propylon contingent consideration, and share-based payments of just under GBP 2 million. The reported tax credit for the year was GBP 1.4 million, resulting in a headline effective tax rate of 11%. However, once we've taken account of all adjusting items, such as exceptionals and amortization of acquired intangibles, the affected adjusted tax rate is 25.6%, up from 25.1% in the first half last year, but in line with the guidance given in December. This adjusted tax rate's increased in the reporting period mainly as a result of an increase in the corporate tax rate in the Czech Republic, one of the group's key operational jurisdictions.
Moving to the next slide, just to show a bridge summarizing the main drivers of change year- on- year to lay out clearly the elements already called out, we've reported lower profit than last year, primarily due to items that can be considered non-trading, most of which were guided to in December. We expect to see improvement in the second half driven by increased revenue, addressing the specific client issues and further realization of our cost reduction program. I'll give a little more detail on the trading performance at a divisional level on the next slide. The non-trading items relate to the sale of PatBase in May 2024, increased amortization, the change in our capitalization approach, which means more technology investment is being expensed in the year rather than being capitalized.
I should add that this is not as a result of a change to the gross R&D investment levels and does not drive any change to cash flow. Finally, FX, primarily being the year-on-year impact of the realized and unrealized hedging programs, with almost GBP 5 million loss this year versus a GBP 4 million gain last year, with a smaller trading FX impact in the period of approximately GBP 2 million. Moving to the next slide and giving some color on the divisional revenue. Language services reported an increase of 4% at constant currency year- on- year. Our data services, AI-led proposition TrainAI, performed strongly with a number of our enterprise clients, and we continue to win new project work in this area.
Whilst we have seen encouraging growth in the core localization services and APAC, we continue to see reduced activity from some of our clients in Americas as they have adapted their priorities to changes in their end markets. The language services first half margin was impacted by some increased competition on price, primarily in the non-West Coast tech part of this division. Adverse mix driven primarily by the accelerated growth of our TrainAI business and APAC core localization, as well as the incremental investment to support the large client volume increases, which was explained at the time of our trading statement. Adverse FX slightly, while our cost reduction programs did broadly offset inflation. In regulated industries, the revenue fell 7% at constant currency.
Whilst we've seen some small growth with our life sciences clients within the core localization services we offer, we have seen reduced activity with our finance and legal clients, and a larger decline in our linguistic validation business in the first half, which we attribute partly to the changes made in our own management late last year and partly to timing variations in client study pipelines. We anticipate a somewhat stronger second half for linguistic validation as these clinical programs advance into new phases. The first half margin for RI was impacted primarily by the reduction in top line revenues and mix changes. Moving to language and content technology, which grew 6% at constant currency year- on- year, driven by a strong performance in Propylon and Language Weaver in particular.
Licensed SaaS revenues grew 12% year- on- year on a reported basis across the portfolio, increasing the SaaS revenues as a percentage of licensed revenues to 43% from 39% last year. The margin in LCT was primarily impacted by the increased expense of technology development and the incremental amortization, the growing SaaS segment, which was partially offset by favorable price as we continue to reflect inflation in contract renewals and ongoing cost reduction efforts. In IP services, finally, we delivered 1% on a constant currency basis, which was in line with our expectations. The increase in revenue included strong performance in renewals and IP research, which have increased significantly year- on- year, offsetting some Eurofile softness. The margin in IP services was primarily impacted by the divestment of PatBase and a change in mix in filing, with a growing dilutive world file versus a declining accretive Eurofile.
There was also a step change in the level of investment in sales capability in the period. Moving to the cash flow, cash generated from operations came in at GBP 46.3 million before tax, which is GBP 37.2 million after tax in line with the same period last year. As can be seen from the bridge, adjusted EBITDA of GBP 38 million is further enhanced by some net working capital improvement in the period. The GBP 14 million improvement has come primarily from trade and other payables, with an increase in accruals, mainly due to timing in items such as the holiday accrual, but also reflecting the focus this year in standardizing vendor terms to 30 days, and the increase in deferred income, reflecting both the increase in SaaS revenues, but also some phasing on the support and maintenance contracts.
The major cash outlays were the final dividend of GBP 37 million, CapEx of GBP 12 million, and tax payments of GBP 9 million. As expected, we received GBP 5 million during the period relating to the final receipt on the sale of PatBase. As a result of all these movements, net debt after loans and borrowings was GBP 27 million. For information, as is typical, lease liabilities of GBP 24 million are not included in these figures. Cash conversion for the year was 171%, an improvement from 30% in the first half last year. This high percentage is skewed by the disproportionate impact of networking capital relative to the half-year adjusted income. On a full year basis, we do expect to see strong improvement year- on- year, but moving back to more normalized levels.
As you can see from the red text on the slide, our operational free cash flow is close to 100% of our adjusted EBITDA. Finally, as a reminder, the group has a $220 million revolving credit facility, which matures on the 6th of August 2027. As of March 31, 2025, we have nearly $100 million undrawn, plus an additional uncommitted accordion of $100 million, providing us the flexibility to make responsible investment and capital allocation decisions without being constrained by debt levels. Moving to the full year outlook, as communicated in April with our trading statement, we expect modest revenue growth in the second half. We also expect to see gross margin expansion as specific ramp-up issues are addressed, and there is further realization of our cost efficiency efforts.
Guidance was provided at the last trading statement on the assumption of a GBP/USD exchange rate of 1.33 for the second half. To better mitigate exchange rate volatility for the remainder of the financial year, the board has now approved increasing hedging to 75% of our net surplus cash flow, which we secured at a rate of 1.35. This, combined with further cost efficiencies within overheads, will deliver adjusted PBT within the range of GBP 60 million-GBP 70 million as guided. I shall hand back to Ben.
Thank you, Candy. Sorry. Thank you. When I joined RWS in January, I did what every new CEO would do, I guess, is talk to a lot of people. I went on to meet many of our employees. They share their ideas, their motivation, but also their frustration.
I met with most of our top clients to understand what they liked about us and where we should improve. I met with many investors who had a lot of suggestions and questions about us. Almost every time, the first question I got was, "Why did you join RWS?" Maybe let me start with answering this very question. The first reason is I'm truly excited about the purpose to unlock global understanding. You know, it's the purpose that fuels international relations, international trade, cultural exchange, and its importance will only intensify. The way we achieve this will evolve, but the need for human-to-human and human-to-machine understanding more and more will only get stronger. Secondly, RWS is a global market leader operating at scale, at a scale that's very difficult to replicate.
Honestly, leading a billion-dollar company, which partners with the largest company on the planet to deliver critical work, is immensely appealing. The blend also of highly skilled service with cutting-edge technology resonates well with my past experience both at Google and at Concentrix. Thirdly, there was a clear recognition from both the board and the management team that RWS needed to transform and accelerate its pivot to a technology-first company. My career has consistently revolved around executing growth strategy at scale. That means articulating the right vision, shaping an effective go-to-market, and building a high-performing team. This is precisely the mandate that I received from the board. There you have it. Three powerful reasons to join RWS: a strong purpose, a powerful platform with global scale, and a mission to change.
On reflection, I was very lucky to look at RWS with a fresh pair of eyes. It's allowed me to discover untapped potential in under-recognized products and to galvanize our team around a refreshed vision of the profound value we deliver to our clients. What I discovered was very exciting. RWS creates enterprise-grade solutions that deliver critical value to the world's largest companies. We have an incredible portfolio of intellectual property and technologies. We operate at a truly global scale, delivering a speed and a quality that is unmatched, as evidenced by our strong NPS scores. Let me show you a few examples of the work we do. We are training LLM models for three of the largest tech companies on the West Coast. We're supporting clinical trials for seven of the top 10 pharmaceutical companies.
We're managing the IP portfolio of some of the most innovative companies in the world, and we support 17 of the top 20 IP law firms. We're enabling data services companies like the London Stock Exchange to create international insights in financial markets. We're supporting the publication of laws in 15 of the U.S. states. All these examples underscore the critical nature of the work we perform for our clients. However, I want to be straightforward, and I think it's important to recognize the issue we face today. Okay, our performance since the merger between RWS and SDL has been subpar. We've lost some momentum, and our top-line revenue has decreased. We haven't fully delivered on the promise of the integration. Today, our current reporting segments are really a legacy of the past acquisitions. Internally, the company has become too siloed, with collaboration hampered by complex processes.
Most critically, RWS has seen a decline in investor trust, mostly because we failed to chart a clear course and constantly deliver on projections. The environment is evolving rapidly. Another way to put what Julie just said is that the pace of innovation and change will never be as slow as it is today. This requires really a shift towards greater speed and agility in our decision-making, in our delivery, and in our development cycles. It has become clear for everyone that we need to change. This is not really incremental improvement we're talking about. It's about strategically repositioning RWS with four objectives in mind. The first one is accelerating growth and increasing profitability. This is really the number one focus to demonstrate the long-term value of RWS. Technology must be elevated and at the forefront of RWS priority.
It's the constant innovation that our customers need, and it also improves the quality of our earnings through a greater share of recurring revenue and high-margin revenue. Increase ownership. We must empower strong leaders across the group with clear responsibilities, and thus fostering a culture of ownership and accountability. Finally, speed. We need to move faster than the competition. We need to be data-driven. We need to constantly improve and accelerate. Agility is key. On that note, I think we haven't been standing still over the last five months. We've already made some really bold moves, and what I'm very pleased is that the team has responded incredibly well to this need to change. In the last few months, we've reorganized all of our product and technology teams under the leadership of Christina Scott.
Christina comes with an impressive track record of leading digital transformation at companies like the Financial Times, News UK, OVO Energy. Most recently, she's worked with a fertility clinic to develop AI products to better patient results. We've articulated revenue growth as the top priority for the company. We've seized projects that couldn't demonstrate clear benefit to that end. We've redefined our product portfolio around three strategic pillars, and we've launched, I hope you noticed, a brand new identity that explicitly reflects RWS Technologies' DNA. Finally, we've initiated an ambitious transformation program to redesign our technology strategy, our delivery system, our go-to-market, and our reporting and insight function. Let me describe now the vision we have for this reorganization and how we think about the various markets in which we operate. At the heart, RWS is a content solutions company.
What we do is we help our clients maximize the value of their most precious assets, their content, their data, their ideas, by doing three things in particular. We help them generate better content at scale. We help them transform their content so that more customers can connect with it. Finally, we help them protect their ideas and their innovation. The content market, as you can imagine, is vast and diverse. Our choice to focus on these three specific services is deliberate. On the next slide, please. We manage them today. They offer immense growth potential, and they represent segments where RWS has a clear competitive advantage. Let's deep dive now in each of those divisions and start with Generate. In our Generate division, RWS provides solutions that enable organizations to produce and publish intelligent content.
We achieve this by seamlessly integrating GenAI tools with robust enterprise-grade publishing software. It is composed of our content technology team and our TrainAI team. Together, they organize, enrich, and manage the publication of content and data. While GenAI empowers content creation at scale, we ensure this creation is supported by structured, authored approaches for critical content types. We work today already with 25% of the U.S. state government, with 35% of the NASDAQ top 20. This division accounts for between 15%-20% of the group revenue through a mix of licensed and professional services. Content creation is exploding at an unprecedented rate. As much content has been created in the last 18 months as in the previous three decades. This content is increasingly personalized, it is multimedia rich, and it is AI-generated.
That creates really complex challenges for companies in workflow control, in brand consistency, and in data security. We provide a range of highly valuable and specialized solutions. Our capabilities in AI enablement span multilingual data annotation, reinforcement learning with human feedback, prompt and instruction tuning, safety and bias testing, and synthetic data validation. By combining linguistic expertise with scalable workflow, we enable LLM builders and enterprise AI teams to improve model accuracy, to reduce hallucination rate, to ensure cultural relevance, and to meet stringent compliance requirements. Our teams really serve as the crucial human layer, teaching AI to understand nuance, context, and real-world relevance, transforming raw algorithms into powerful, reliable, and inclusive AI solutions. We then support knowledge workers in creating critical documents, from engineers drafting product specifications to government staff drafting new laws to global organizations providing service instruction across jurisdictions.
Our customers range from EV manufacturers, software creators, national government, to leading audit firms. We have launched also new AI-powered functionalities to increase the productivity of those knowledge workers through smart search, summarization, and first draft creation, for example. I am just going to take a glass of water. Sorry. As an example, you have here an engagement that we have with a leading U.S. West Coast tech company. It illustrates the scale and depth of the operations that RWS can deploy. In that case, it involves 5,000 data scientists, data specialists engaged in sophisticated prompt engineering, meticulously testing and validating the output to fine-tune a technical chatbot. This level of precision is critical even for the most advanced AI company. The market in which GenAI operates is vast and growing at a solid double-digit rate to an estimated $10 billion by 2028.
You see there are two different markets. The way we see it is that in the content management space, on the next slide, our suite of products positions us strongly to compete against what are generally bespoke and expensive solutions. Our client retention is very high, and our margin is significantly higher than the group average. We believe that our product innovation with the AI feature I mentioned makes our software even more valuable and will lead to expanding the market. Our TrainAI division, which has seen explosive growth, is playing in the bigger data annotation market. We're winning deals against well-known operators, and we believe more companies will need this type of service to train and adapt LLMs to their specific industry. If the margins are below our group average, in that case, the nature of the work is by definition very manual.
We've invested in specific systems to increase our productivity. Content value is really realized through its reach. Our transform division is dedicated to making content accessible and engaging to wider audiences. The transform division is our largest by revenue, and it consolidates now our state-of-the-art linguistic AI technologies with the full breadth of our localization services that we had across the company. We handle the transformation and enrichment of assets for over half of the Global Fortune 100 companies. This ranges from software localization to website, contracts, video, and audio. We also test the content in the network of labs that we have around the world and are able to recreate any technical environment. The revenue comes as a mix of subscription, professional services, and a good portion of work, which is still priced per word. The language industry has enabled global trade and international relations.
Those relations are only increasing in volume and complexity. We see this, for example, in the growth rate of our APAC business or in the increasing demand for complex languages. The industry is undergoing rapid transformation, profoundly impacted by AI. The overall market size is substantial at around GBP 24 billion and growing slightly, but the growth hides a significant shift within the market from a service-first solution to a technology-first solution. The reason RWS has been so resilient compared to peers is because we own the leading translation operating system for the industry. Our strategic shift is to embrace a technology-first solution for our clients and to lead with the very product that disrupted the industry. Unlike consumer-grade solutions you may all be familiar with, enterprise language solutions are far more complex.
While minor errors in B2C scenarios can be tolerated, mistakes in an enterprise context can lead to severe consequences: brand damage, regulatory non-compliance, and significant financial losses. Research, for example, from Forrester indicates that poor localization can result in a 20%-60% decrease in customer satisfaction. Our software, Trados, is already the software of choice for the industry, from 300,000 freelancers to large enterprises to even competing LSPs. Trados has been the clear benchmark of language management. We've invested to offer a full cloud-based solution with natively integrated Language Weaver, and we've added a suite of AI functionality to keep translation and workflow more efficient. I'm delighted to announce today that RWS has acquired the IP of Papercup to integrate into Trados and offer full media capabilities from dubbing to subtitling.
The acquisition strategically positions us in the high-growth AI dubbing market, estimated at around $800 million in 2024, and part of the larger dubbing market, which is estimated at $4 billion. Papercup's focus on high-quality human-in-the-loop dubbing perfectly aligns with the demand for both scale and quality in this exploding multimedia content space. They have worked over the years to perfect what they call the prosody of a video, which means the emotion, the intonation, and the tone of voice, which is crucial for high-quality broadcast. As a matter of fact, they have worked with premium video rights owners. Many of its dubbed videos have actually been broadcast on TV, where the bar for quality is much higher. As part of the deal, RWS has acquired all IP, software, AI models, and data generated from inception.
Now, from a language management standpoint, we see really three options for companies to manage their language capabilities. A SaaS option where the software is integrated into their tech stack, and that's ideal for most corporate, for internal comps. This is, for example, how the London Stock Exchange is using Language Weaver to ingest and translate RNSCs from around the world. Then you have managed service, where enterprise outsource to RWS the management of the content workflow within the tool. That's ideal for large brands publishing high volume of content in over 40 languages. This is how we work with Dell, for example, who decided to move to a single vendor solution. Finally, professional services, where we provision language specialists for specific needs, such as linguistic validation, post-edit, user experience testing. That's ideal for life science, software tests, legal validation.
That is really the nature of the work we do for all of our pharmaceutical clients. Our vision is that a language platform integrates into the technology stack of a company to natively provide this multilingual capability to all other systems. Not any linguistic capability, but one that is customized to the need, to the tone of voice, and to the particular glossary of each company, including the list of complex acronyms that we all have. It connects with a CRM, with a CMS, with a software development tool, with a productivity suite. All of this while keeping your data private and secure. ServiceNow, for example, one of our key accounts, has moved from the definition of software readiness to include a full localization. Language cannot be an afterthought. It is rooted in the design.
Canva, as another of our key accounts, has taken the same bold bet from the start of the company. They wanted to be global and available in 100 languages. Canva is an incredibly successful AUD 40 billion unicorn from Australia that is disrupting the creative world through powerful AI solutions. Rather than tell you, let me show you in their own words the value that RWS is bringing to them. [audio plays Canva representative highlighting Canva's global reach of over 220 million monthly users, it's partnership with RWS since 2016 to deliver culturally relevant and localized experiences]
[audio plays about Canva representatives describing how Canva expanded to 100 languages, emphasizing RWS's adaptability and support in major events like Canva Create and Droptober, and highlighting RWS's role in providing innovation, collaboration, and localization expertise]
[audio plays about Canva representative emphasizing that over 100 million users access Canva in non-English languages and highlighting RWS as an adaptable extension of Canva's team that consistently delivers]
[audio plays continuation of Canva representative] You may be familiar with the term LSP, a language service provider. Slator, one of the analysts in the industry, has now split the market between service provider and technology provider, really, and they're called LSI and LTPs. Within the overall $24 billion market, the language service integrator, LSI, would deliver managed outcomes, capturing still the majority of the buyer spend, about 87%.
This confirms actually that buyers, particularly large companies, continue to demand managed solutions and outcome ownership from their providers. They need a partner capable of end-to-end responsibility to deliver assured language outcomes. However, the language technology platform, LTPs, is focused purely on technology and is the fastest growing segment. It accounts for roughly 13% of the market, but growing at a CAGR of around 15%. Our strategy positions RWS to capture both the demand for managed solutions, where LSI are strong, and the high-growth technology segments, where LTPs are strong, converging best of both worlds. With this new organization, we expect to find additional efficiency in the service side of the business and focus our sales on the strategic segments and clients. This will give us margin in line with the group average. Leading with technology solutions, we expect to shift faster our revenue to a licensed model.
Having a large base of existing customers gives us a clear advantage to push and recommend our technology suite. The margin in this part is a lot more attractive too, and we aim to grow our customer base beyond the large enterprise into the corporate segment with a dedicated online sales team. We also expect to benefit from the faster-growing multimedia segment, which combines technology and human-in-the-loop for higher quality output. Finally, our Protect division. The Protect division represents the foundational heritage of RWS. It is helping innovators protect their ideas around the world. In fact, RWS has proudly supported IP professionals since 1958. Today, it accounts for 15%-20% of our revenue and encompasses a broad range of services to cover the full patent lifecycle. RWS is a well-respected brand in this industry where we work, as I say, with 17 of the top 20 patent filers.
As part of the business, localized patents are still priced per word, but increasingly, this is a service business with a growing share of annuity revenue. We are a species of innovators. The worldwide spend on research and development reached an astounding $2.8 trillion in 2023, which has translated over the years into 18 million patents filed worldwide, patents that need to be monetized, renewed, sometimes recorded to a new owner. In 2023 alone, there have been 3.5 million patent applications, which represented an increase of 7.6% over the previous year. In that context, corporate IP departments face increasing pressure to drive IP revenue, while IP law firms are challenged to demonstrate value in diverse ways. Against this backdrop, and with the advent of AI, traditional IP outsourcing models and IP management solutions must evolve. We have kept innovating to bring automation in this highly manual and meticulous industry.
Today, we extend our support across the entire IP lifecycle, in particular through our unique crowd approach to patent search, leveraging an unrivaled network of 40,000 crowd resources across more than 160 jurisdictions, often finding information in obscure books in not-so-well-known libraries. We have also completed the strategic expansion into IP portfolio-based solutions, including patent renewal and IP recording, like Candy mentioned, and it is showing incredible progress already. Our ability to tailor and localize IP solutions in high-growth markets like China, South Korea, and Japan makes us an attractive global partner in Asia, alongside our traditional markets in Europe and North America. For example, we have just recently signed an agreement to manage the international patent portfolio of Alibaba, a symbol of innovation in China.
Becoming the partner of choice for managing patents of our customers promises a significantly higher quality of earnings characterized by an increased share of recurring revenue. The IP market is vast, at about GBP 2 billion, and remains growing steadily, and the majority of the market is based on annuity. I'm particularly enthusiastic about this division, where RWS enjoys an outstanding reputation. We see substantial opportunity to expand our market share in a segment that's accurate in margin profile. Our refreshed sales team also has already been successful in cross-selling our new product to our existing clients. None of this transformation would be possible without a strong value foundation. Peter Drucker famously said, "Culture is strategy for breakfast." I've already witnessed at RWS how nurturing a strong culture has helped us accelerate our transformation.
The enthusiastic response to my weekly message to all employees, the positive feedback on our rebrand, has been incredibly encouraging. We're doing significant work to anchor those values in the culture across the company and to define what it means to be part of RWS. At RWS, we partner; it means we prioritize our clients, we work collaboratively, we engage openly, we win together. We pioneer; it means that we innovate relentlessly, we lead with ambition, and we never hesitate to disrupt ourselves. We progress; we set very high standards, we learn continuously, and we grow through decisive action. Finally, most importantly, we deliver; we own our outcomes without excuses, we move with intent, and we build unwavering trust.
This strategy, altogether, will really elevate the nature of our relationship with large enterprises, and it will allow us to become a content solution provider for a larger set of clients. Operationally, RWS will be structured around three agile business units led by an accountable leader and supported by a lean and efficient support function, which will allow us to move faster and innovate further. Each division will benefit from a central technology team, attracting the best talent and incorporating AI swiftly without compromising on privacy, security, and quality. Our revenue model will become clearer and more predictable as it shifts towards a higher proportion of SaaS revenue or fixed-based FTE services. In terms of timing expectation, I'm discovering the magic of committing to the market. I've committed in February to articulate our strategy today, and this has been really a catalyst internally to move fast.
In the next 100 days until the end of our financial year, we are focused on two things: delivering our 2025 results with a particular focus on maximizing our revenue and securing our margin. We're not waiting to implement changes that improve our model immediately. Secondly, finalizing our organization, the process, and fine-tuning our business plan. By October 1, we will begin to operate under the new structure: generate, transform, and protect, each with clear accountability, a tailored go-to-market, and aligned investment priorities. When we meet next at our full-year result in December, we will share our midterm guidance and our new operating KPIs. We are crystal clear about the work ahead, but we are energized by the opportunity. This is about execution, discipline, and building the platform for sustainable growth.
I trust this presentation provides you a clearer understanding of RWS's formidable capabilities and the opportunity that lies before us. I'm a strong advocate, I hope you've noticed, about simplification. In the world of language, there is a timeless quote attributed to Blaise Pascal, which goes, "I didn't have time to write a short letter, so I wrote a long one instead." Allow me to leave you with a very simple, concise summary of where RWS is heading. You will have three agile divisions delivering strategic content solutions focused on growth and quality of earnings, unified by a state-of-the-art technology and a strong, thriving company culture. Thank you very much. We can move to Q&A. James.
Thank you. It's James Ward from Deutsche Numis. I have three questions, please.
Firstly, a lot of what you've said in that strategy update seems to center around the business historically maybe being slow to execute and adapt on strategic and technological changes in its markets. Can you give a few anecdotes of examples of where the company has been slow to execute over recent periods?
I'll give you a simple and anecdotal example, but we have assets that sit in various segments currently. For teams to collaborate, it has become increasingly complex. They have to recharge each other. They do not know if it is the priority of the segment or the next segment. I think having a simple way to operate, first bringing all of the technology together, is absolutely key. Otherwise, you have competing teams doing similar works and not collaborating efficiently. This whole process of recharge, etc., has made us too stiff.
That's really what we want to implement. You have now three divisions that have a clear go-to-market. They have a clear market to operate in, and they're really crystal clear about what their priorities are.
Thank you. Second question. In recent years, there's been a lot of talk about the implementation of the LXD model within RWS. That appears to be sort of notable by its absence in this presentation today. I just wondered if that's deliberate or whether that still remains a sort of a relevant element of the strategy going forward.
LXD is a platform that enables RWS to benefit from its scale when it comes to delivering linguistic services. Now that we're bringing all of the linguistic services under the same roof in our transformed division, LXD is very focused on improving the delivery speed and efficiency of those segments.
I think we need to act as the leader of the market. We need to show to our clients that we have access to the best resources at a price that's attractive. More importantly, by operating under the same roof, we also have a technical suite that allows us to have better quality, better speed in the service that we deliver. LXD is continuing. LXD is transforming. LXD is transforming to have also a technology-first approach and see how a lot of the work in the localization market is around management of projects. It seems not straightforward, but when you receive thousands of documents from a company to translate in 40 languages and following some specific workflows for compliance reasons, the management of those projects is very complex.
Still, we believe that in many, many cases, we can automate some of those workflows through new agentic AI capabilities.
Thank you. The third question, and I appreciate you're giving more context around this at the prelims in December, but there are aspects of what you said today that would appear to suggest that the business requires sort of further capital and operating investment. Could you just maybe give a little bit more color around that and any indications on return on investment?
Yeah, we're finalizing all of the investment profile, but I guess what I can share with you today in terms of philosophy is we're not on a spending spree. There's as much savings and efficiency that we will find to be able to fund our investment.
We're very, very, I think, responsible in the way we will spend money and focusing also on tools, projects, technology that actually fuel this strategy in each of those divisions. There are some projects that are more comfort-based. We're not touching those.
Thank you.
Katie, go on.
Hopefully, you can hear me. Why are you paying a dividend instead of keeping the funds to invest in the turnaround?
I think we're paying a dividend to tell investors to stay with us and wait and see. Trust us on this strategy. We have a clear capital allocation policy, and so I cannot talk about the dividend at the end of the year. The message today is to come with us on this journey.
Thank you. One more online. For how long will the West Coast tech companies need you to train LLMs? And what happens when that's done?
I appreciate the million-dollar question. It's a very complex one. When ChatGPT came out three years ago, we thought it was done. They've never spent more CapEx than today. We were invited three weeks ago at a strategic supplier day from Alphabet, who shared that just this year, they're going to spend $75 billion in CapEx. That's just Google. Our view is that the need for refinement and understanding will continue to stay. A stat that always blew my mind is, I think, on the 3 billion searches that Google receives every day, about half of them they've never seen before. We need, it's not about just understanding the history, which is, okay, maybe done, but it's understanding how the world is changing. I don't know how you woke up this morning, but there's a ton of new news that never crossed our mind before.
How we react to those news, how we make sure that the LLM doesn't develop a weird strategy about the world, I think, is constant.
Cheers. Tom Cowan from Investec. Just a couple, please. Firstly, just picking up on something that James touched on before in terms of LXD. I can't remember exactly how many people are employed within that specific platform to sort of run that platform. But based on your comments, Ben, does that point to potential headcount rationalization across LXD? Don't want to sort of preempt stuff that might come out in October, but just trying to get a flavor for that.
Just secondly, in terms of, I think, again, this is probably going to come out in the KPIs, but in terms of the sort of medium-term targets, from what you've seen thus far from being in the business, do you think this business can get back to sort of margin levels that we've seen, maybe that we were seeing maybe a couple of years ago? Or do you think structurally the market is in a different place now and we should expect a sort of lower run rate margin moving forward? Thanks.
LXD currently employs roughly 1,500 in-house translators. We're constantly reviewing the benefit of having a translator in-house versus outsourced. And believe me, I get challenged every day on that number. There's two elements that we need to take into account.
First of all, having in-house translators, if you manage to occupy them at a certain level, not only are they cheaper than using freelancers, but they also guarantee a level of quality for the biggest clients that creates a lot of stickiness. Our team managing some of the West Coast companies, they know the glossary, they know the tone of voice. I would see even a world where some companies pay us on an FTE basis to manage those linguists on behalf of those companies. There is this element. The second element is on some languages, it gives us also a high competitive advantage because there is in the world maybe a pool of 20 maximum people that are able to do, I do not know, mechanical in Finnish. That is a selling argument. It is more anecdotal.
We are constantly reviewing the productivity and the benefit of having an in-house linguist. That is not just for LXD. It is for across the company, reviewing or operating a model. I talked about leaner, more agile. Of course, that is making everybody accountable. On the margin level, not going to lie, there is a lot of pressure from the market. People want to have more and more content translated. Anecdotally, we have received in the month of May more content than we have ever received before in our LXD division. There is a need for our service, but our clients need this to be more efficient, cheaper, etc. I think the margin will improve if and when we actually move to this technology first and we are able to really automatize and leverage the technology that we own. We know it is possible. We have demonstrated on many clients.
It's more a matter of replicating it and installing this operating model.
So, pick you up on that final point. You sort of say it will improve. Will it improve from where we are at the moment, or do you think it will improve versus where we've been a couple of years ago? Just trying to sort of be clear on that.
Technically, if it improves, it improves against everything. Will it get better? It's too early to say. One of the complexities I acknowledge in those three divisions is you have really three different businesses operating with three different business models, three different billing mechanisms. I would hope you can wait a bit on this and understand more the mechanics of those three divisions. James.
Morning. James Bayliss from Berenberg. Pardon me. Two questions, if I may. The Papercup acquisition.
I guess, obviously, you've outlined a significant organic undertaking to reposition the strategy of this business. Should we be thinking of Papercup as more opportunistic, or would there be expectation we could see other M&A at the same time as this transformational change at RWS? Question two, just around, I guess, the employees and the stakeholders within the business. You've consulted with them quite heavily as to what has been problematic in the past and where they need to change. Can you talk about the journey you'll take them on to get them behind this new strategy? Thanks.
Thank you. Papercup. Papercup is a startup that was based here in London. They have raised a significant amount of capital over the years, and they're really focused on high-quality dubbing.
You have the pure AI dubbing, and a famous name in that market is 11 Labs, for example. You have the full manual dubbing. There is no space in the middle. That is really what Papercup has been focusing on. Their technology has been recognized by some really premium brands, as you have seen in the slide. They have struggled to invest in a commercial model. What they did is they decided to carve out their engineering team. They had an amazing engineering and data science team that went to Scale AI. We were able opportunistically to get all of the other assets for ourselves. This has been opportunistic in the sense that it was a great acquisition, I think, for us.
It is also strategic because I think we are looking for quite some time on how to tap into this multimedia market. Going forward, I have announced the arrival of Joseph Ayala, who is with us today. Joseph is leading our corporate development, and we are looking at opportunities that can help bring us some new capabilities, potentially, especially in technology for each of those divisions, or some gaps in certain markets. We will be extremely strategic and responsible in our M&A, if any. When it comes to employee, in one of the first meetings I went to, I went to visit our team in Brno in Czech Republic, which is still one of the biggest centers that RWS is. I was doing this all hands with all the team.
One of the employees, in front of everybody, asked me, "Sorry, my variable pay is 5% of my base. How do you think I'm motivated to go above and beyond?" I thought it was courageous. In front of everybody, I asked him, "Would you be okay to lower your base and increase significantly your variable?" He said, "Any day, tomorrow." People were agreeing to this. I think that's really, in the company, there's a willingness to change. There's a willingness to win. I think that's what makes me very proud with the employees. The employees have responded incredibly well with the new brand, which is fresh, modern, and really represents the assets that RWS has. Tomorrow, we have RWS Day.
The same day across the world, we're celebrating RWS with engagement for all of the employees and all of our operating centers. I'll be sharing more of the strategy, more of the organization. We've increased significantly the level of communication with all the employees. I think I really see this enthusiasm perpetrated around the company.
One question from online, which is, have you identified any particular growth opportunities in which you want to invest in product areas and so on?
We have. We have very much. No, jokes apart. We've been working, especially over the last months, to identify for each of the divisions the strategic growth levers. We have clear ideas on each of the divisions of where are the biggest growth opportunities. As we build our budget for the future years, we're focusing on those growth pillars.
I do not really want to detail those immediately, but we have done a lot of the work for each pillar.
One last one, as I can see on the webinar, brief, Candy. The presentation did not suggest a target return on capital employed under the new strategy. What do you think is realistic in the medium term? My contention would be that current ROCI is far too low.
Thank you for the question. I think the fairest thing to say is that we continue to look at the strategy to enhance all our key metrics, and we would look to see an enhancement on the ROCI as well as the earnings and growth KPIs. More to come.
That is it. Thank you very much. Thank you for joining us.