—highlights before passing to Justin. FY 2025 was a year of meaningful progress for Appen. At the group level, we delivered $230.8 million in revenue, up 4.5% on FY 2024 when you exclude the impact of Google. Growth was driven predominantly by new project wins and expansions in Generative AI. Generative AI is a clear growth driver for the market and a positive signal that we are executing well against our strategy. On profitability, we delivered $12.2 million in underlying EBITDA, excluding FX. The full year margin was 5.3%, with a strong end in year, with Q4 coming in at 18.2% EBITDA margin. Gross margins are moving in the right direction, driven by wins in higher value Generative AI work. We also continue to capture operational efficiencies through technology, innovation, and automation.
Looking at our two segments. On Appen Global, the full year revenue was $127.9 million, which was down year-over-year. We ended the year on a high note, with Q4 revenue of $41.4 million, up 56% on Q3, and EBITDA of $10.2 million at a 24.6% margin. The Q4 growth was predominantly driven by new project wins, including a $10 million Generative AI opportunity that grew faster than we anticipated and has carried into FY 2026. We continue to focus on the operational turnaround, talent is a major part of the shift we are making across Appen Global. In the last 12 months, we have added over 20 experts to the team, coming from either customers or direct competitors.
Appen China had an exceptional year, delivering $102.9 million in revenue, up 75% year-on-year, with EBITDA up 640% to $10.6 million. Growth was driven by both new and expanding Generative AI-related projects, the momentum heading into FY 2026 is strong. Across the group, 44.1% of Q4 revenue came from Generative AI, up from 34.8% a year ago. Finally, we closed the year with $59.8 million in cash on hand. We continue to drive operational efficiencies across the business. In FY 2025, we realized $10 million in annualized cost efficiencies through technology and automation. In summary, FY 2025 was a year that demonstrated real progress, revenue growth, improving margins, exceptional performance from Appen China, strong momentum as we head into FY 2026.
We've built a stronger team, and we're winning in the right parts of the market, and we have the balance sheet to continue executing on our strategy. I'm proud of what the team has delivered, and I'm confident in our trajectory. With that, I'll hand over to Justin, who will take you through our financials in more detail.
Thank you, Ryan, and good morning, everyone. A reminder that we report in U.S. dollars and that all comparisons are to the year ended 31 December 2024, unless stated otherwise. Starting with the FY 2025 profit and loss snapshot on slide five. As Ryan has already mentioned, revenue increased 4.5% to $230.8 million. This excludes the impact of Google. Within our operating segments, Appen China revenue grew by 74.8% to $102.9 million, with Appen Global down 21.1% to $127.9 million. Appen Global delivered a strong finish to the year, which was up 56% quarter-on-quarter. Growth in both segments is predominantly driven by new and expanding Generative AI projects.
Gross margin improved by 100 basis points to 40.3%, with the improvement driven by a greater mix of Generative AI projects. Underlying EBITDA, before FX, grew 251% to $12.2 million. The increase was driven by gross margin improvement, as well as the $10 million cost efficiencies achieved via technology and automation. The cost out is net of talent upgrades. I won't talk to slide six, as we've covered this data, so over to Appen Global revenue and EBITDA on slide seven. The chart on the left shows quarterly revenue, and on the right-hand side it is underlying EBITDA. The charts show strong momentum in Q4 for Appen Global. The Q4 growth was driven by new project wins and includes the $10+ million GenAI opportunity we previously announced.
Gross margin improvement was achieved due to the high priority GenAI projects. In addition to the improved gross margins, the $10 million annualized cost efficiencies were achieved within the Appen Global segment. As I just mentioned, the $10 million is net of talent upgrades, who, along with the continued focus on data quality, were a vital factor in winning and delivering new projects. It was a strong finish to the year, with good momentum going into FY 2026. The full year was impacted by lower volumes than expected in Q1 to Q3. Full year revenue was $127.9 million, down 21% excluding the impact of Google, and underlying EBITDA was $5.8 million, down from $9.2 million in FY 2024.
Over to slide eight, which shows quarterly revenue and underlying EBITDA for Appen China and reflects the strong market position Appen China continues to hold. Revenue grew each quarter, with Appen China achieving $102.9 million in revenue for FY 2025, which was 75% growth on FY 2024. Growth was driven by new and expanding LLM-related projects. Appen China exited the year with annualized revenue exceeding $135 million. Pleasingly, in addition to revenue growth, profitability improved throughout the year, with improved gross margins due to a greater mix of GenAI projects and increased revenue from high-margin pre-built datasets. Appen China is also capturing scaling efficiencies due to tight OpEx controls as revenue expands. Turning to slide nine for the profit and loss summary. I won't talk to all the line items.
However, there are a few additional points to highlight. The decrease in employee expenses reflects the benefit of cost out efficiencies achieved in the Appen Global segment. It was partially offset by some incremental OpEx in Appen China to enable China's delivery of the strong revenue growth. Underlying NPAT improvement was minimal, despite the EBITDA improvement, due to an increase in non-cash amortization. Statutory NPAT was impacted by an additional $5 million acceleration of non-cash amortization in relation to acquired platforms. I'll finish up with the cash flow summary on slide 10. The cash balance at the end of the period was $59.8 million, up $5 million from December 2024. The Australian dollar equivalent of the cash balance is AUD 89.5 million. Cash flow from operations improved by $23.4 million- $22.4 million.
The year was positively impacted by the receipt of a payment from a major customer in the first week of January versus December 2024 as scheduled. However, adjusting for this still results in approximately 100% conversion of EBITDA to cash flow from operations. Cash flow used in investing activities was $3.5 million higher compared to FY 2024, due to a slightly higher investment in product development and new facilities for the Appen China business. Cash used in finance activities of $4.9 million reflects lease payments. Cash was used to fund operations and CapEx. That concludes the financial performance slides. I'll now hand back to Ryan.
Thanks, Justin. I'll now provide an update about the market opportunity and our strategy. I'll start on page 12 and explain why we're confident about the market that we're operating in. There are three structural trends that are driving demand for the kind of work that Appen does, and importantly, all three are linked to major economic drivers. First is the globalization of consumer AI. As AI products are deployed across different countries and cultures, you need human data that reflects the cultural nuance. This is also directly linked to the digital advertising market. As AI powers more personalized, localized advertising experiences across global platforms, the demand for culturally and linguistically accurate data grows with it. That creates a sustained need for the kind of multilingual, multicultural data that sits at the core of what Appen has been doing for a very long time.
The second is enterprise AI adoption. McKinsey estimates $2.6 trillion-$4.4 trillion of economic value is at stake across enterprise use cases. As companies move from experimentation to deployment, they need AI models that understand their industry. That requires expert-generated data. Layered on top of that, we're seeing real momentum in agentic AI. These are the systems that don't just answer questions, they take action autonomously. The model builders we work with are focusing heavily on agentic AI right now. It's still very early, moving very quickly. These labs are actively experimenting, figuring out what good looks like. What's exciting for Appen is that we're in the room with them as they experiment. We're working closely alongside some of the leading AI labs as they explore and build out these capabilities.
We have high expectation that this work will continue to be evolved and be a major growth driver for Appen. The third is new form factors, particularly humanoid and industrial robotics. Models that interact with the physical world need data grounded in real-world examples. This is an area where we're already active in, with projects covering robotic simulation and real-world data collection. The work we're doing now positions us well as this market develops and investment in robotics continues to grow. Moving on to page 13. The market opportunity is clear, but what matters is whether we're actually winning projects in the right areas. This slide gives you a concrete view of where we're competing and where we're winning. These are examples of projects that either started in Q4 FY 2025 or are commencing in Q1 FY 2026. A few things stand out.
We're landing projects with new frontier Lab customers. That's meaningful because these relationships tend to expand over time. We're also expanding within existing relationships, winning new data modalities and new capabilities, like coding with customers that we already work with. The previously announced $10 million-plus opportunity is worth highlighting. It grew faster than we anticipated, driven by the quality of our delivery. The customer expanded the project specifically because we exceeded their quality goals. This is the kind of outcome that builds long-term relationships, and it's currently ongoing with potential for continued expansion. We're also moving into more advanced projects, including domain-specific coding, robotic data simulation, multi-speaker voice model training. These aren't commodity data tasks. They require specialist capabilities, and that's where Appen is increasingly competing and winning. On to page 14.
Winning the kind of work described on the prior page requires the right people with the right technical expertise, and over the last 12 months, we've made deliberate investments in our team. We've added more than 20 industry experts across go-to-market, operations, and workforce management. On the go-to-market side, Brian Jenkins leads our sales team, bringing experience from Scale AI and Snorkel, and a track record in enterprise sales at Salesforce and Oracle. On the operations side, Jeanine Sinanan- Singh brings a background in AI research from Surge AI and Microsoft, with a computer science degree from Harvard. Francisco Rivera leads workforce operations, with experience scaling complex marketplace operations at Uber and Angi. We continue to grow the capability of the team to complement our historical strengths at Appen. Moving on to page 15.
Underpinning our ability to deliver high-quality data to our clients is an asset that is difficult to replicate: our global workforce. This is no longer just a generic crowdsourced workforce. It's increasingly a verified pool of domain experts in fields ranging from computer science and law to finance, health science, and engineering. When a customer needs domain-specific expertise at scale and with global reach, there are very few organizations that can deliver it. That's a structural advantage, and it deepens as our workforce grows and our domain coverage expands. Bringing this together on page 16, I want to be clear about why we think Appen is structurally well-positioned. It comes down to four things. First, the next wave of AI data is squarely in our area of expertise.
The shift towards generative AI, multimodal models, and enterprise AI deployment requires exactly the type of complex, domain-specific, human-generated data that Appen specializes in. Second is our track record. 30 years of delivering complex, large-scale data projects for the world's leading technology companies is not something that a newer competitor can replicate quickly. Particularly as the demand for AI globalization grows, our experiences operating across languages, cultures, and regions at scale is a genuine differentiator. Third, our position in China. Appen China's revenue is now significantly larger than our nearest established local competitor. We're the top vendor of choice for Chinese technology companies and model builders, and as that market continues to grow, we are very well-placed to grow with it. Fourth is technology. Our platforms and processes are built to scale and adapt with customer requirements.
Technology is not just an efficiency lever for us, it's what allows us to take on more complex workflows, deliver them consistently, and continue to improve our cost base. I'll now describe our 2026 focus areas and provide an outlook statement. On to page 18. Our FY 2026 priorities are focused and deliberate. There are four areas where we believe execution will drive the most value. Data quality is first, and it sits above everything else. The reason customers choose Appen and expand with us is because we deliver quality. That doesn't happen by accident. It requires consistent investment in our processes, technology, and people. Second is customer growth. We're directing our go-to-market effort towards the segment with the highest spend potential, hyperscalers and foundation model builders.
These are the organizations driving the next wave of AI investment, and deepening those relationships is where we see the greatest opportunity. Third is new data segments. We're actively co-innovating with our customers to expand into new data modalities and techniques. Working closely with our customers ensures we have close visibility into new growth vectors. Fourth is operational efficiency. We'll continue to drive technology-led efficiencies across the business, building on the $10 million in annualized cost savings we achieved in FY 2025. We head into FY 2026 with strong Q4 momentum, a clear strategy, and a balance sheet that gives us the flexibility to execute. We're focused on building our trajectory and continuing to grow our position in the market. I'll close on page 19 with an outlook statement.
As shared throughout the presentation, we remain highly confident on the AI data market and the potential for Appen to meaningfully contribute to the development of leading foundation models. We continue to see positive signals on LLM-related growth, including from Appen Global and Appen China customers. Tight cost controls remain in place, in keeping with the company's focus on managing costs in line with the revenue opportunity. As in previous years, Appen Global revenue continues to be mostly derived from project-based work, and seasonality skews revenue to H2. Considering this, we provide the following FY 2026 group guidance: revenue of $270 million-$300 million, an underlying EBITDA before FX margin of 5%-10%. That finalizes our presentation today. I'll now hand back to Sam, who will moderate the Q&A.
Great. Thanks very much, Ryan, and thank you, Justin. As a reminder, research analysts can raise their hand via Zoom should they wish to ask a verbal question today, and the rest of the audience can submit written questions via the Q&A function, at the bottom of your screen. Our first question comes from Josh Kannourakis at Barrenjoey. Josh, please unmute your line and go ahead.
Great. Thanks, Sam. Good day, Ryan and Justin. Thanks for taking my questions. First question, just with regard to talent, it's obviously been a big focus for you guys, and you've capitalized on some of the disruption at some of the other peers in the market. That obviously translated through to some good deals that have happened towards the end of last year. Can we talk about into 2027, as you think to the pipeline, how much of these talent hires really opened up the pipeline and some of the customer relationships you've been able to do?
Yeah, thanks, Josh. Good question. You know, the talent that we've brought into the business has enabled us to do far more technical work with our customers. What it means is that we're really at the bleeding edge of the work, and a lot of that work is very early with the customers. How the LLM development cycles work at the moment, there's a lot of experimentation. When we see a model technique that has a positive impact, those projects can expand really, really quickly.
What we're seeing at the moment is a lot more activity across the business in terms of the projects that we're doing, you know, particularly in some of the more advanced types like we laid out in the presentation, and, you know, we're really confident that those early relationships we're building with the researchers, getting in on the ground floor on these new modeling techniques, will evolve into much, much larger projects over time.
Perfect. Then just with regard to guidance, like, obviously, you have stayed away in more recent times from doing it, so bringing a guidance range out there suggests a heightened level of confidence in this year. Can you talk us through a little bit more just around, you know, what needs to happen at maybe the upper and lower ends of that guidance? If there is continued development in some of these new customer relationships, what that means in terms of where there's upside risk to that as well.
Yeah, sure. Look, we are feeling more confident, and there's a few factors there. One is, you know, as China becomes a larger portion of our revenue, that is, you know, less variable to quarter-on-quarter shifts, because in China, our customers come to us and, you know, ask more for an allocation of resources rather than a specific project base, so that gives us a lot more confidence in putting guidance.
In terms of the swings, we just need to be, you know, executing well against the projects that we have and making sure that we're catching the next wave of growth in the market that comes through these, you know, like I said earlier, more innovative projects in areas that, you know, we know that there's market opportunity that we need to be breaking into. With the team that we've got, with the, you know, projects that we're seeing at the moment, we feel really confident about the opportunity.
Got it. Final one from me, just on the ad side of the business, like, obviously, that's going to be a big focus, OpenAI specifically have noted that—
Yeah
— you guys have a lot of specific expertise. Can we just flesh that out? I think it's worth fleshing out a little bit more around exactly how you would play into that and what you think that ecosystem might end up looking like, and effectively, where you end up being able to be in that value chain moving forward.
Yeah, we're super bullish on, you know, continuing to support the digital advertising market. You know, there's two factors to it. Right now, you know, my take is that there's going to be a pretty strong arms race in major U.S.-based technology companies. You think Meta and Google, who have the dominant position in the market, but also OpenAI and X, formerly known as Twitter, as an outsider. There's two real things that these companies are going to be focusing on. One is building great consumer experiences of products, so people are spending their time on the platforms, and the second is building out a really effective digital advertising ecosystem, which makes it easier for advertisers to place ads on the systems, but they're also more effective and targeted.
This is very similar to the, to the search and ad relevance that Appen has been doing for a very long time. What becomes really important to get these models to work, both on the experience side and the digital advertising side, is making sure that they work at a global scale. What many of kind of the engineers and researchers would consider works well in the West Coast of the U.S. is very different to what's going to happen in Pakistan and India, in Southeast Asia. Getting the models to work at a global scale requires a lot of feedback about cultural nuance and specific training to make sure that the models are working in those geographical areas.
Appen's got a tremendous amount of strength there, and we're already seeing some customers really lean in, into that, so we're super bullish on this next phase of the digital advertising arms race.
That's right.
Great. Thank you very much, Josh. Next question comes from Conor O'Prey at Canaccord. Conor, please unmute your line and go ahead.
Morning, Ryan. Just, maybe just a question on the, how the, how you see the unit economics stacking up for this year. China business, strong revenue trajectory, the ex-China, the global business, less so. Would you expect a lower gross margin this year? With a sort of 23% revenue growth, approximately at the, at the midpoint of the range, do you need to invest in the, in the cost base, the OpEx base to support that growth?
I'll start with the second one. We don't see any requirement for any investment in the OpEx base, to support the growth, so we're really confident in our OpEx base at the moment. In terms of the gross margin, look, it will play out on the mix between China and non-China, but also the gross margins that we see within the global business unit. You know, we're pretty comfortable when you look at the FY 2025 gross margins, that that's something that we can achieve in FY 2026.
Thank you. Thanks.
Thanks very much, Conor. Next question, we've got a couple coming through, around the China business. Firstly, can you just elaborate on what's driven this growth, please, and specifically, what type of customer?
Yeah. Good question. We work with almost all of the major technology companies in China, which includes companies that are building the foundation models and incorporating them into their applications. A big driver of growth has been the development and the implementation of Generative AI models on two horizons. One is for the domestic China use, but it's also for Chinese companies that have international operations, where we're helping them expand through data that represents the international markets that they operate in. Very heavily focused on Generative AI, both for domestic applications in China and the international expansion.
Great, thank you. Just in terms of concentration, specifically within the Appen China business, can you elaborate on that at all with respect to top customers?
I can talk at a high level. You know, it's a reasonably diversified mix of customers that make up our Appen China revenue. Like I said, we work with many of the large players over there, so we're very comfortable with the composition of revenue, and we'll continue to look to drive revenue from those customers.
Thank you. With respect to, I think, the relative valuation you've pointed out against a domestic peer, what's the company doing to realize this value disconnect?
We point out the peer in terms of the relative market share that we're getting. We're just continuing to focus on operating the business. We think we've got a great asset in China. There's a great market opportunity, so it's heads down on focusing on the potential for China.
Just one final question on China. When you think about the performance of the China business versus the, I guess, the contrast this financial year with the Appen Global business, how do you think about allocation of resources into, you know, the Appen Global business versus Appen China?
We do look at it a divisional standpoint, and we look at the needs of both of the business units and what we consider appropriate investment, and we make our decisions appropriately. We've been investing in growing China for a while now. We're starting to see the payoff now, which is great, and we look to get operational leverage. You know, China, sorry, Appen Global, we have a as seen through last year, a bit more of a lumpy profile, which is driven by the project nature, but we're confident at the spend levels that we've got at the moment, that we can deliver meaningful growth in Appen Global.
Okay, great. Thank you. Just on seasonality, can you elaborate on historical seasonality, including first half, second half SKU, both on a revenue and EBITDA perspective? In relation to EBITDA, what specific initiatives have you got in place to drive EBITDA margins through the balance of FY 2026?
Yes. Some of the EBITDA margin, we are focused on revenue growth in the business. As I said, we don't need to add any cost to the business from an OpEx standpoint. Clearly, our crowd cost flex up and down, but we, you know, feel really strong about the operating leverage that we can get in the business, which will drive EBITDA as revenue grows. In terms of the first half, second half SKU, traditionally, as a business, we have seen a SKU towards the second half. Typically, what we see in January and February from some of our customers that run a calendar year planning cycle, that there is a bit of replanning that can occur, and that's kind of one of the main factors that drives the SKU.
Great, thank you. Next question is coming from Jennifer at Jefferies. Jennifer, please unmute your line and go ahead.
Thank you, team. Thank you, Ryan. Just a couple from me. Probably, you already touched on some already. The first one, it's still about the SKU. When you talk about it's traditionally, in line with previous, but how do we look at the global and China, two segments individually?
Yeah, China's a little bit harder to call out because it has been driving the right direction. I would think of it more of an Appen Global SKU, where we see the mix between... We see more of a H2 being a bit heavier than H1.
All right. On the margins, could you also give us a bit more color on China margin going forward, like FY 2026 and beyond, given the customers and the product mix dynamics?
Like I said, we're starting to see some operating leverage in the China business, which is really good, and we see that through the kind of improving EBITDA lines. We feel really confident about kind of that margin trend continuing as revenue grows.
All right. That's all from me. Thank you.
Thanks, Jennifer.
Thanks, Jennifer. Next question, just on cash balance. Do you expect your cash balance to continue to grow further by the end of this calendar and financial year?
I'll take that one, Ryan. The cash on the balance sheet is there to support the working capital requirements. Obviously, it's a decent amount of working capital required as we grow quickly, that's because we pay the crowd ahead of getting customer receipts in the door. There is some working capital needed. Obviously, timing impacts it, too, when we're looking at it as at date, at the end of the year or whatever. As we've kind of called out, we're very confident in the year, we're driving towards profitability, naturally, you would expect the cash balance to go up over time.
Great. Thank you. How is Appen adapting to the rise of synthetic data and automation? Specifically, is there any second-order impact of these observations or trends?
Great question. Synthetic data is a really important part of the market, and we're incorporating that into a lot of the projects that we're delivering for our customers, where there's a mix of synthetic data and human data. You know, it really comes through the solutioning that we're bringing, which is far more, as I called out in the presentation, you know, far more consultative, and, you know, much more tech-led in the solutioning. We're definitely incorporating all of the techniques that we can to deliver data for our customers that are high quality, good unit economics, which definitely includes synthetic data.
Thank you. Maybe just one, time for one more question, more around the longer term. What areas does Appen see the biggest growth opportunities over the next few years? How do you think about long-term margins for the business once fully scaled and automated?
I think, I'll refer back to the presentation, where we called out three areas that we see really significant growth potential. First is the B2C market, which is fueled by the digital advertising spend. We think there's huge opportunity for the globalization of consumer-based AI models. Secondly, is the enterprise focus with a real steer towards agentic AI. We are seeing, you know, agentic AI has been around for a while, but we're seeing huge focus from the foundation model builders on domain and expert-specific agentic AI models. We think that's a real area of growth. New form factors like robotics, like new language models, AR/VR.
I think there's many things that, you know, will continue to evolve in the AI market that we don't have great visibility of at the moment, which is, you know, really exciting for us as we look towards the future. In terms of margin, we're looking to build, a high-growth business with sustainable margins and supporting our customers, delivering high quality. You know, we've put out a longer-term, you know, end of FY 2027, target of 10%, so I'll leave that point out there now. But over the longer period, again, just driving a healthy business is the focus.
Okay, great. Thank you. I think that's all the time we have for questions today. If you do have any follow-ups, please feel free to email them through, we'll endeavor to come back to you. Maybe with that, Ryan, I'll just pass it back to yourself and/or Justin for any closing comments.
Yeah, thanks, Sam. Thanks, everyone, for joining the call today. We really appreciate the support. I'll just summarize FY 2025. Look, it was a year where we demonstrated real progress at Appen and, you know, strong momentum into heading into FY 2026. We're highly confident on the AI data market and the potential to meaningfully contribute to the development of leading foundation models. We've built a stronger team, and we're winning in the right parts of the market, which gives us huge confidence in, you know, both FY 2026 and beyond. I'm super proud of what the team has delivered and really confident in our trajectory. Thanks again for your support. We really appreciate it, and looking forward to seeing you at our next market update.
Great. Thank you very much for joining today's Appen FY 2026 results call. Thank you, and enjoy the rest of your day. Goodbye.