Appen Ltd
ASX:APX
US |
Johnson & Johnson
NYSE:JNJ
|
Pharmaceuticals
|
|
US |
Berkshire Hathaway Inc
NYSE:BRK.A
|
Financial Services
|
|
US |
Bank of America Corp
NYSE:BAC
|
Banking
|
|
US |
Mastercard Inc
NYSE:MA
|
Technology
|
|
US |
UnitedHealth Group Inc
NYSE:UNH
|
Health Care
|
|
US |
Exxon Mobil Corp
NYSE:XOM
|
Energy
|
|
US |
Pfizer Inc
NYSE:PFE
|
Pharmaceuticals
|
|
US |
Palantir Technologies Inc
NYSE:PLTR
|
Technology
|
|
US |
Nike Inc
NYSE:NKE
|
Textiles, Apparel & Luxury Goods
|
|
US |
Visa Inc
NYSE:V
|
Technology
|
|
CN |
Alibaba Group Holding Ltd
NYSE:BABA
|
Retail
|
|
US |
3M Co
NYSE:MMM
|
Industrial Conglomerates
|
|
US |
JPMorgan Chase & Co
NYSE:JPM
|
Banking
|
|
US |
Coca-Cola Co
NYSE:KO
|
Beverages
|
|
US |
Walmart Inc
NYSE:WMT
|
Retail
|
|
US |
Verizon Communications Inc
NYSE:VZ
|
Telecommunication
|
Utilize notes to systematically review your investment decisions. By reflecting on past outcomes, you can discern effective strategies and identify those that underperformed. This continuous feedback loop enables you to adapt and refine your approach, optimizing for future success.
Each note serves as a learning point, offering insights into your decision-making processes. Over time, you'll accumulate a personalized database of knowledge, enhancing your ability to make informed decisions quickly and effectively.
With a comprehensive record of your investment history at your fingertips, you can compare current opportunities against past experiences. This not only bolsters your confidence but also ensures that each decision is grounded in a well-documented rationale.
Do you really want to delete this note?
This action cannot be undone.
52 Week Range |
0.265
2.87
|
Price Target |
|
We'll email you a reminder when the closing price reaches AUD.
Choose the stock you wish to monitor with a price alert.
Johnson & Johnson
NYSE:JNJ
|
US | |
Berkshire Hathaway Inc
NYSE:BRK.A
|
US | |
Bank of America Corp
NYSE:BAC
|
US | |
Mastercard Inc
NYSE:MA
|
US | |
UnitedHealth Group Inc
NYSE:UNH
|
US | |
Exxon Mobil Corp
NYSE:XOM
|
US | |
Pfizer Inc
NYSE:PFE
|
US | |
Palantir Technologies Inc
NYSE:PLTR
|
US | |
Nike Inc
NYSE:NKE
|
US | |
Visa Inc
NYSE:V
|
US | |
Alibaba Group Holding Ltd
NYSE:BABA
|
CN | |
3M Co
NYSE:MMM
|
US | |
JPMorgan Chase & Co
NYSE:JPM
|
US | |
Coca-Cola Co
NYSE:KO
|
US | |
Walmart Inc
NYSE:WMT
|
US | |
Verizon Communications Inc
NYSE:VZ
|
US |
This alert will be permanently deleted.
Earnings Call Analysis
Summary
Q4-2022
Appen reported a disappointing FY 2022, with total revenue falling 13.1% to $388.5 million, largely due to reduced spending by global customers amid weaker digital advertising demand. Despite these challenges, the New Markets segment grew 15%, driven by a 36% increase in revenue from China. The company is focused on establishing greater operational rigor to enhance sales and profitability, aiming for $10 million in annual cost savings starting in H2 2023. While the first half of FY 2023 is expected to be challenging, underlying EBITDA is projected to be materially lower than the previous year. Detailed strategic direction will be provided in May 2023.
Thank you for standing by, and welcome to the Appen Limited FY '22 Results Release [Operator Instructions].
I would now like to hand the conference over to Mr. Armughan Ahmad, CEO and President. Please go ahead.
Thank you, Melanie. Good morning, good afternoon and good evening, everybody. This is my first call with Appen. I look forward to speaking with all of you. I really appreciate the opportunity to join Appen.
Today, I'm here with Kevin Levine, our Chief Financial Officer; and Rosalie Duff, our Head of Investor Relations. We're looking forward to having a brief update with all of you, and there's an agenda slide on Slide 3, where we will cover 3 sections for the presentation today. First, Kevin will present our FY '22 financial performance. After Kevin, I'll spend some time sharing with you why I joined Appen. My initial focus areas, my vision at Appen and information on our FY '23 outlook. We will then go into Q&A.
Before I hand over to Kevin, I would just like to say and take a moment to share that it's my absolute pleasure to be here at Appen and leading Appen with this incredible team we have here. and that I'm very much looking forward to meeting many of you in person over the coming days, while I'm in Sydney and Melbourne.
Now I'll turn it over to Kevin for the financial updates. Kevin?
Thanks, Armughan, and good morning, everyone. Thank you for joining the call this morning. Before we move to presenting the FY '22 financial performance, I just wanted to say that I'm very excited that Armughan has joined Appen. We are working closely together to improve transparency and accountability. My team and I are working closely with the business units to improve our commercially focused FP&A partnership to help drive improved performance.
Starting with the financial summary on Slide 5. Total revenue decreased 13.1% to $388.5 million. This primarily reflects the reduced spend by some of our Global customers on core programs with high margins. Our Global customers are reporting lower revenue and earnings driven by weaker digital advertising demand. This also flows through to our New Markets business, where we saw a decrease in Global Product revenue.
Excluding Global Product revenue, the New Markets business recorded revenue up 15% on the prior corresponding period with China being the key driver. Underlying EBITDA and associated margins were significantly impacted by lower-than-expected revenue as well as higher cost base to support China and Quadrant growth and investment in product technology and transformation.
We report an underlying net loss of $22.8 million due to the decrease in EBITDA and increased amortization of product development spend as well as for intangibles acquired as part of the Quadrant acquisition. We recorded a statutory net loss after tax of $239.1 million after booking a noncash impairment charge of $204.3 million. No final dividend has been declared.
While our FY '22 performance was disappointing, there were some highlights, which we'll call out on Slide 6. In Global, we reported revenue growth in our second largest client of 20% year-on-year. In terms of New Markets, we achieved the following: 184 new clients, 25 of non-global deals over 250,000, Enterprise increased their average deal size from 61,000 to 140,000, China recorded strong revenue performance, with revenue up 36% year-on-year, and we remain the leading data AI company in China, servicing 9 out of the 10 major auto clients. Our overall customer NPS was 22 and above our internal targets.
Turning to the revenue performance on Slide 7. At the group level, revenue was down 13.1%, and the majority of this decline can be attributed to Global Services as well as related impact from Global Products. We did see improvement in the second half, with second half group revenue up 12% on the first half and Q4 revenue up 11% on Q3. Global Services revenue declined 13%, which reflects challenging market conditions that led to a reduction in some of our large core projects. We did see strength in some segments with a 20% increase in revenue from our second largest clients.
We saw improvement in the second half as Global Services revenue was up 18% on the first half. Q4 revenue was also stronger, up 12% on Q3. New Markets revenue declined 13.8% and was impacted by lower contribution from Global Products. Excluding Global Products, New Markets grew 15% year-on-year. And this was underpinned by 36% revenue growth in China. Excluding Global Products, New Markets for growth in the second half with revenue up 4% in the first half and Q4 revenue up 14% compared to Q3.
Turning to Slide 8. Group underlying EBITDA before the impact of FX decreased 82.8% to $13.6 million due to lower revenue, lower gross margin impacted by customer project mix and investment in people and OpEx to support growth in China and Quadrant as well as product technology and transformation. Both revenue and cost increased 12% compared to the first half. However, we saw Q4 costs declined 9% on Q3 due to cost-saving initiatives, while revenue increased by 11%.
The Global Services division reported EBITDA of $54.5 million, down 40.2% in the prior year. This reflects the impact of reduced customer spend on revenue and gross margin, as previously mentioned. In addition, there was higher delivery overhead from increased project activity and a higher allocation of indirect costs resulting from investment in product engineering and transformation costs.
Global Services costs grew 15% in the second half compared to the first half, while revenue grew 18% in the same period. And Q4 costs compared to Q3 reduced 3% while revenue grew 14%. New Markets reported an EBITDA loss of $36.5 million compared to an EBITDA loss of $11.5 million in the prior year. This reflects a reduction in Global Product revenue and gross margin.
Gross margin reduction in some of the business units impacted by customer project and service mix, higher costs in China to support growth, full year of quadrant trading with additional investment in costs to drive future growth and higher allocation of indirect cost to Enterprise resulting from investment in product engineering and transformation. Excluding Global Products, the second half rate of cost increase slowed through the half. While second half costs grew 10% against revenue growth of 4%, Q4 costs reduced 3% on Q3 costs, while revenue was up 14% for the same period.
On Slide 9, we present the balance sheet. Our balance sheet reflects no debt. The reduction in total assets was primarily due to the decrease in noncurrent assets from the impairment charge of $204.3 million. This impacted goodwill and certain intangible assets in the New Markets, excluding China, group of cash-generating units as announced on 13 February. Please refer to the appendix for more detail on impairments.
In terms of current assets, we saw a decrease in receivables due to lower revenue volumes. Our receivables exclude some work done in December as time-based billing milestones were not met. This is reflected in the increase in contract assets, and this work will be invoiced in January. Total liabilities decreased by $3.4 million, mainly due to a reduction in our leased office footprint. Noncurrent liabilities include the Quadrant earn-out liability of $19.1 million, payable in cash all shares in 2024.
Subsequent to year-end, we have received credit approval for the extension of our debt facilities to 3 January 2024. This extension agreement is expected to be finalized in the next few weeks. Our full debt refinance process is planned for the second half of FY '23 to achieve an appropriate debt structure to support the current and future needs of the business. Our existing debt facility has been resized to a working cap facility of $20 million to reflect the current needs of the business.
Over on to Slide 10, which shows our investment in product development. This year, investment totaled $41.2 million and reflected our continued focus on product development and customer experience. We invested in a new roster fill initiative for Global, automation of crowd labeling services, build various new machine learning models and build new functionality for crowd fraud detection in Appen Connect. 50% of the spend was capitalized, reflecting a greater focus on the current product set and platform stability. The FY '22 product investment represented 10.6% of revenue, and we expect to invest a similar percentage of revenue in FY '23.
Over to Slide 11. Our cash balance decreased in line with lower revenue, lower margin, higher OpEx and CapEx spend and a minority investment in Mindtech. Strong cash conversion continues to be a key strength of the business. We saw cash conversion from EBITDA increased from 77% to 135%, aided by positive working capital movements.
Operating cash flow was impacted by lower volumes and higher payments, and this was somewhat offset by positive working capital movements and lower tax payments. Cash was used to pay OpEx, dividends, CapEx and investments to support future growth.
Thank you. That concludes the financial slides, and I will now hand back to Armughan.
Thank you, Kevin, for sharing the financial numbers. It is a pleasure to be speaking with all of you today, as I mentioned earlier, and this is my first investor meeting at Appen. I'm going to cover 4 topics in my presentation today. It will take about 20 minutes. So I hope you bear with me, and then we will turn it over to Q&A.
First, I'll share with you why I joined Appen and why I'm excited about the business. Second, I'm going to discuss my immediate priorities for the business. Clearly, the results that Kevin just shared are disappointing. My top priority is to establish greater operational rigor to accelerate innovation, drive sales and deliver profitable growth back for Appen and our shareholders. Third, I will spend some time talking about growth opportunities for Appen. This will include some exciting product announcements related to generative AI and large language machine models that we announced at ASX and the broader market today. Finally, I will provide some commentary about our FY '23 outlook, and let's move on to Slide 13.
AI is a dynamic and fast-growing market and Appen as a category leader. We power many of the world's leading artificial intelligence models run by some of the most advanced companies on the planet. The clients on this page are only a subset of our clients, which are also including biggest technology companies and Enterprise brands in the world.
Since joining Appen on January 9 of this year, I've been fortunate to meet 2 dozen of our clients across United States, Europe, Asian markets and Australian markets. I've been here in Sydney for the last 1 month and have met not only clients, but our system integrators as well as other officials, just trying to understand the market, and it's been fascinating to hear how much respect Appen has in the Australian market.
In addition to our brands listed here, we work with 8 of the 10 largest tech companies in the world, and we're continuing to add more clients every year. In FY '22, as Kevin mentioned, we added 184 new clients, which is amazing to see. We have also have had long-standing client relationships, which has also have been really, really nice to see that many of our clients rely on Appen, and they have been Appen loyal clients for over 9 years.
Our client list is very impressive, but what has impressed me the most is the level of innovation we support with our clients with our Appen people all around the world and how we're delivering unique customer experiences for our clients to deliver new customer experiences. We are thrilled to be partnering with our clients to create industry-defining products at a very large scale.
And over the last 25 years, Appen has evolved from a total constant change and has adopted this constant change that we've been in since our beginnings in Linguistics, the speech and language, through to supporting ad and search models to image recognition. And now we're working on state-of-the-art generative AI models. Our ability to adapt and thrive in a market of constant change is a key strength to Appen, and it positions us very well to continue the evolution and drive maximize -- and drive and maximize the key transactions that we want to drive going forward.
Appen, if you move to the Slide 14, Appen is an AI platform services business. Now that we see so many of our clients that I just shared with you, we provide AI platform services to most of the big tech clients on the planet and the data and services we provide are used to improve the core business operations, often directly linked to billions of dollars in revenue.
Our Fortune 500 clients value our expertise greatly. And we have gained a lot of insights by working with the big tech clients and many of the C-suite clients I've met in all my travels, the Chief Data Chief Technology and Chief Information Officers with some of the top Fortune 500 clients, they have told me that they appreciate Appen's consultative approach, and they would like us to share our best practices around scaling AI models for Enterprise-wide adoption and what we have learned from big tech clients. Our AI data platform made up of our [indiscernible] and AC products that Appen producers underpins all our services today, and we have advanced platform to manage our crowd operations and a leading annotation platform that is directly used by our clients and our internal data.
So when I think of Appen and why I joined Appen, I certainly don't think of happen as a BPO or business process outsourcer. There's nothing wrong with being a BPO, but we are much more than that, and it's a fast-growing AI business, and it's just getting started. We are an AI platform services company. We combine our platform and expert services to partner with our clients to help them deliver world-class AI models and for them to deliver new customer experiences.
If you move to Slide 15, I will take you through my immediate focus areas. So I just shared the 2 reasons why I joined Appen, and now let me tell you the work ahead of us together to make sure that we totally realize that the FY '22 results are disappointing and we have much more room to improve. I have successfully led many businesses in the past in successful turnarounds back to growth. And in my past, I have come to realize that an operational rigor is much needed to ensure that we scale a business that's very quickly gone from tens of millions to hundreds of millions. And as we now look to grow Appen to its next phase of profitable growth again. Number one, cost management and reigniting growth is my immediate focus. As you will see on Slide 16, we have listed 6 priorities, which I'll share with you in a lot more detail. But first, as I walk you through on Slide 15, first and foremost, I'm focusing on this operational rigor across our busines,s, and I'll go through a lot more detail on the next slide.
My next priority is also increasing our product velocity to match our investments in the product and Engineering group that Kevin just talked about. We will drive much more impact from our investment in technology, including new client-facing products that we launched today, automating our labeling process with generative AI that we announced today and driving internal productivity improvements to ensure that our percentage of on-time Engineering delivery is improving. We will also continue to build a world-class go-to-market function.
A big part of growth vision is to build a consultative selling approach by creating more standardization in our sales and go-to-market motion. We will also elevate our marketing and brand awareness and bring some North American style swagger to Appen. You will see us expanding our partnership ecosystem, which is critical to reach more and more clients around the world and build industry verticalized solutions by leveraging our learning from the big tech advanced AI industry.
Finally, we will deliver AI for good. This is a very important to our clients, but also it's a core value to us at Appen.
Now on to Slide 16. As we discussed, establishing operational rigor across the business is my top priority to drive positive momentum and take Appen back towards a profitable growth trajectory. Appen has strong fundamental in our high-growth markets, and I see those strong fundamentals. But there are some very important items that require immediate attention. Number one, we are establishing a business management system to track and measure financial and operational performance. The business management system will inform a weekly, monthly, quarterly business review cycle to all revenue-generating business units. We are building a measurement process for product and Engineering to increase on-time delivery and track the benefit of every feature that we released.
We are reviewing our costs in the business through a zero-based budgeting exercise. This is something I've done in many of the organizations before, zero-based budgeting essentially requires all of our business unit functions and all of the business functions that exist within Appen to get back to running their operations in a very effective and cost-effective way. They have to prove back to me why we need to spend a dollar for every business unit and benchmark it against what the industry norms look like, so that we know that we are fit for every business unit. It doesn't matter if it's a back-office business unit or a front office or middle office, we are going to drive this zero-based budgeting model.
We are also creating a target operating model for delivery functions. I have come to see that our delivery functions are not optimized for a shared delivery function. We have a lot of silos that we need to work towards improving -- towards a new target operating model that has shared efficiencies that we can drive.
And then finally, we will be utilizing generative AI to assist the automation of our annotation process. We have launched our first product today that's related to driving automation and generative AI into our own offerings, and you will see more and more of that. And the list of activities on this page are my top priority, and I'll share an update with you. I'm a big believer in say-do ratio. So what you will see in May is what I'm telling you today, I will provide you a progress update on that. And then I would like now -- I would now like to share a bit more around the market and the approach for growth.
If you move to Slide 17, AI is changing the world, and I see a huge opportunity for Appen to play in this world. It's a $16 trillion market, going very fast to -- by 2030. And some of the industry leaders who are talking about AI, that all of you read in the newspaper or online or your social media of choice every day is people like Sundar Pichai, the CEO of Alphabet, the parent company of Google. And he is famous in declaring that AI is one of the most important things humanity is working on, and it's more profound than electricity and fire. Satya Nadella, the CEO of Microsoft has made a great code -- has set a great code to where AI is just the beginning of its S-curve, and the near-term and long-term opportunities are enormous. Mark Zuckerberg of Meta has declared that one of his goals is for Meta to become a leader in generative AI and how they're incorporating that in their metaverse platforms. These are just a small set of leading companies that are investing heavily in AI, and I fundamentally believe that AI is a huge opportunity that's going to continue to grow that Appen needs to be part of.
If you move to Slide 18 on how we are living in this world of exponential change and we need to keep up with this level of change. Just to give you a data point here, ChatGPT has revolutionized how humans interact with machines, and it's created a new experience that's unlike what we have seen in the software industry in the past 50 years. When we compare ChatGPT to other technology companies that have delivered new experiences, the vast difference in the rate of adoption is just astonishing.
For example, it took 16 years to reach 100 million mobile users, 7 years for the Internet, 4.5 years for Facebook, 2.5 years for Instagram and even TikTok took 9 months. And ChatGPT only took 2 months to get to 100 million users. Just as a point of reference for everybody, I joined Appen 2 months ago. So in that time since joining AI product has gone from 0 to 100 million users. We are in this world of exponential change and Appen will also move very fast to keep up with this exponential rate of change with our clients and our people.
Let me move on to Slide 19 of generative AI and the opportunity I see for Appen. First of all, I want to do some myth busting, does Appen play in generative AI? We already play a huge role in generative AI, and we have been playing that role for the last 6 months. Let me explain. ChatGPT and other recent breakthrough AI applications are built on technology called generative AI, and it presents a significant opportunity for Appen.
The majority of large AI models until recently have utilized deep learning approaches, and Appen has been very successful in supporting clients and -- on data collection, data annotation and relevance work to get their data ready for AI. That is what we have been doing for a long time, and we're very good at it.
In particular, we lead the market in providing relevance work in this deep learning area that you see on the left and our clients rely on human feedback. And some of our biggest top tech Global brands rely on us. 76% of our revenue in FY '22 was from relevant services. We are also working with generative AI solutions with our clients for the past year, and many of our current projects are underway.
The strong capabilities we have built in deep learning by capturing human feedback is exactly what's required in support of generative AI. This is a technique that's called reinforcement learning with human feedback, also known as RLHF, reinforcement learning with human feedback. We're excited to formally announce this RLHF solution, the human -- the reinforcement learning and human feedback solution based on generative AI products.
Today, we have also announced that on the ASX, and you will see that announcement hit the U.S. and the Global markets starting Monday, their time and in addition to 2 other products. So generative AI presents a significant growth opportunity for Appen, and we are perfectly suited to serve that market because we have been doing that work already that requires human in the loop and now with generative AI, our clients have already engaged with us.
Moving on to Slide 20. From an Enterprise perspective, our clients are using AI to create new customer experiences. But as I've gone to see many of the Fortune 500 clients, that I've met over the last few months, many of them are telling me that they are having some major issues on getting data ready for AI and how they can apply that data for their AI markets. Majority of the Enterprises are telling me that they have some common friction points for adoption for AI solution. First, they have lots of unused data, up to 80% of their data is unstructured, and it is not being used. That's number one. Number two, their data preparation is very difficult. Studies have shown that 80% of large Enterprises data scientists are spending 80% of their time preparing data for AI. That is what Appen does really well. Third, data is required to build a new customer experiences. It's often uncomplete, new and external data is often required. That is, again, what Appen does really well in our data collection capability.
And then finally, we see that many Enterprises do not have data infrastructure required to support production levels AI. And that is something that Appen again, does very well. But a lot of Enterprises are looking to us to say, how have we done that work for some of the most advanced and most matured big tech data companies who have already structured a lot of their data to drive insights out of it, and that becomes a competitive advantage for many of the Fortune 500 clients. If you're in financial services or consumer retail or health care or government, that's what they're all asking for. And the good validation point is that Appen has done that work, and we are now going to start providing that to our clients.
If you move to Slide 22, this will provide you a bit more understanding of high-growth products and me being in the tech industry and tech and services industry for the last 27 years, I've seen this S-curve many times before. And I believe that Appen's also going through our S-curve.
Let me try to explain what I mean by that and try to introduce the concept of S-curve, which many of you may likely already be familiar with. High growth technology companies, as I said, that have product offerings follow a very similar pattern, and there are many such examples. Typically, after their launch, it takes a while for adoption to kick in. When it does, it enters a period of very, very high rapid growth. And after a time, it slows down. And we're thinking about growth in terms of performance of S-curves. And Appen's performance has track to an S-curve to the T. To deliver performance, we need to maximize our current S-curve. And for the next level of growth, we need to maximize our first S-curve to fund our second S-curve and we need to identify and execute against that S-curve.
Before I share our thoughts on how Appen feels our second S-curve needs to be, I'd like to provide you an example of a company that has done really well there. If you go to Slide 23, we see an example of how Microsoft has stacked S-curves to delivery and sustained growth that they have provided. They have started with Windows as their first S-curve and they got to almost 100% market share but then that first S-curve got flattened out for almost a decade for them under -- when Bill Gates moved to Steve Ballmer as the CEO, and you saw that there was a bit of that flattening of the S-curve. Satya Nadella, the new CEO came in, and really turned it from a know-it-all to a much more of a learn-it-all organization. And their second S-curve was Microsoft Azure, and they were able to grow that, and now they have got a $2 trillion market cap.
Many of the other companies I worked at have also had to find their second S-curve because in technology industry we need to start optimizing for it. And now Microsoft has found their third S-curve, and they already started investing ahead of time. In 2019, they invested in OpenAI, the creator of ChatGPT and now they're deploying AI into their products. And what they announced a few weeks ago, where they have already announced AI into their search product to compete against Google Search.
Let me move on to Slide 23, which is for me to talk to you about Appen's S-curve. So on Slide 23, you'll see that we've been thinking a lot about how to evolve our business in light of massive AI opportunity. Now I want to make sure that all of you are very clear, we still see a huge opportunity in our first S-curve. We have been successfully helping our clients get their data ready and grow tremendously on the success of the quality data and leading big tech companies, and that has been Appen's first S-curve. This will continue to be a major focus as we move upmarket in the big tech clients, which we call our Global clients as well as our New Markets in our Enterprise, China and other areas. And we see a lot of growth remaining in our core market, which is our S-1. In our next S-curve, we expand beyond data for AI to applied AI and build industry vertical AI solutions to help our clients adapt AI for their industry use cases. We see -- we have seen firsthand the possibilities of AI and also the changes -- the challenges facing many of our Enterprise companies trying to adapt AI and deliver new customer experiences.
Our industry-focused solutions will bring together our AI platform services with other leading AI companies and services companies. This is what I call our ecosystem approach to deliver business outcomes for large Enterprises. We call this our S-Curve 2 for Appen. And then eventually for S-curve 3, if we execute S-curve 2 well, we will be able to productize and reuse and scale many of the solutions that benefit the solutions in our second S-curve and provide a stand-alone recurring revenue value for our clients. The focus will be on business-to-business, B2B point solutions that solves major pain points for our clients, where they are currently operating in a very expensive do-it-yourself. If you look at financial services, health care, consumer retail, there's a lot of very expensive DIY platforms that require efficiency. And they are already wanting to use AI to drive the efficiency.
If you move to Slide 24, this brings our growth vision together for one -- for Appen on one page. Let me start from the top down. I call this my house slide. If you actually look at the top of this growth vision, it shows you how our clients are looking to use AI to enable customer experiences for their own customers.
However, as I explained, some of the challenges that they're facing today is they have a lot of unused data, they have preparation -- data preparation is very difficult for them, data is very incomplete for them, and Enterprise data pipelines are very nascent. If you see in this house picture slide that on the left, that is where Appen has been focused on. Appen has been focused on the deep learning side where we're getting data for AI life cycle. And many of the companies have overcome these challenges, but they now are looking for Appen for support. So how will we provide support?
First, we will continue to provide software and services to support data for AI development. To date, we have focused on getting the data ready for deep learning, and we will continue to innovate and grow there. As you may have already seen today, we are announcing 3 products on the generative AI. And you can see in our press release and blog post for more details on it. But at a high level, these are the products related to building large language models and using technology to create training data sets. Our teams have been working feverishly for the last few months to get these products out. And I am so proud of our product engineering and delivery teams that they were able to get that done. A great effort by all of them. So I really congratulate them on that.
The second layer you'll see is the Appen AI platform that develop the industry verticalized solutions. We have been redefining our position in the AI market with valuable training data, and we are looking to partner in market to deliver a solution with an ecosystem. And we have recently launched solutions focused on automotive industry that you may have seen. And you will see more and more of verticals such as financial services, consumer, retail, industrial markets as well as health care. But building AI cannot be done in a silo, and it requires an ecosystem of partners. I just came from a very large big 4 consulting company and after spending many years in big product companies, this continues to be a challenge where services, consultative services. It's not about the what, it's about the how and how do you actually use AI is very important than many of the big ecosystem partners or system integrators and software vendors and cloud hyperscalers are already working with us, and we are going to accelerate that adoption even more.
If you go to Slide 25, if we do this really well, which we will, expanding our solutions will significantly expand our total addressable market. This is a chart from IDC and a research analyst firm, and it estimates that AI market will reach $15 billion for 2026. That is for the AI for data life cycle market. But when you look at the services market, with AI platform services, it will now reach to $300 billion. That's almost a 20x increase.
And we have a clear right to win in the expanded TAM, the total addressable market and we expand -- as we expand our core market into higher order AI platform services product. A subset of the big tech clients are already at the forefront of AI innovation, which I mentioned that 8 out of the top 10 technology companies by market cap already work with us. We know what great looks like. We have a lot of institutional knowledge, a lot of data to build impactful and trustworthy artificial intelligence models. Most of the C-suite leaders that I've met over the last few months have been asking me, hey, what are some of the best practices that you have seen and what data platforms have you seen Armughan, that you can bring to us and how we can move fast to help them adopt automate repetitive functions to improve our customer experiences, and we can certainly help them.
Finally, I will also mention that how do we really win in this market. I would say to expand our TAM, we already are working with many industry leaders, and we have knowledge on how we're working with industry verticals with the expertise of artificial intelligence to help them deliver this market. And we have very close relationships with many of the major AI infrastructure and software providers. We've already done this work for them, but we need to really move up market and have the relationships, which I have and a lot of the leadership that we have here has those relationships with the ecosystem partners, and we're excited about the opportunity to not just do their back-office work, but actually move into the front office by partnering with them. Microsoft, Amazon, we already have partnerships with, and you will see us expand many more of those partnerships.
We also have the right team, as I mentioned, to deliver our product engineering teams, our sales go-to-market team, our delivery teams. And what you will see from me in this new leadership is that we need to set up operational metrics and a rhythm to inspect what we expect. I live by that mantra. My team has very quickly understanding that I am -- I've got a personality where I'm pleased but not satisfied and currently, certainly not satisfied with the performance that Appen has provided, and we will look to change that. And we're going to be very clear on our vision, and I'm confident that the ability to capture significant value during this transition for Appen is going to be critical to us, and we will share a lot more in our May Technology Day that is structured.
If you go to the Slide 26, this is very important to me and very near and dear to my heart, which is how do we create a positive impact. I'm a huge believer in leading with purpose first, ensuring that, that purpose has perspective on why we are doing this and then making sure that leads to our ability to create a lot of value for our shareholders. The work we do is very impactful and how we do that work is it provides a positive impact on society. As you've seen in recent news, there has been a lot of hallucination problems with AI that its requiring reinforcement learning and human feedback to improve it. Hallucination essentially means that there is a lot of inaccurate information that AI is sending back to humans. But because it sounds really smart, it sounds like a hallucination problem. So that is what needs to be fixed. Regardless of generative AI solution we offer, we at Appen want to be a continued leader in building AI with the responsibility by design framework to build ethical and trusted AI.
We have over 1 million plus crowd model -- crowd people that work with us, and that crowd model for us provides opportunities for all abilities and backgrounds for people to work with Appen. We are proud of our diverse crowd, which spans many cultures, ethnicities, age, life stages and occupations and our crowd code of ethics defines our dedication to the well-being of this crowd. At Appen, we thrive to do good, be good, lead good. That is what I'm calling our AI for good strategy. Today, our trust and safety leader, Samantha Chan; and General Counsel, Carl Middlehurst, have published a blog on our point of view for AI for good. I would encourage all of you to check it out.
By doing good our society is going to be better for it where our crowd model and social impact initiatives as, for example, the work we have done for Larrakia National Aboriginal Corporation of people in Australia, where our goal is to preserve the Larrakia language by using our AI solutions. Number two, which is how do we be good. We'd be good by doing the right things internally. As an example, our Crowd Code of Ethics is a good example of that, and we have become a signatory to the United Nations Global Compact organization. We are an ASX-listed company. We are very responsible. We're not privately listed VC backed like some of our competitors. We are doing the right things, and we are making sure that we're reporting on that.
And then finally, the lead good by working with our clients to improve how AI impacts the world. We're a leading partner with wormed Economic Forum on defining best practices for inclusive AI. This was talked about at Davos, by the way. Again, we need to really start talking about what Appen does with great swagger, and there's a study -- case study that you can check out on our website on how we're working with the World Economic Forum. We strongly believe that the ethical approach to AI will be a competitive differentiator moving forward, aligning our social and financial goals as a business.
Finally, I want to move to our 2023 outlook. I hope that this explains our vision about the growth prospects for Appen and how I plan to drive an operational rigor at Appen that aligns to our profitable growth and ambitions.
I would like to now talk about this on Slide 27, where we have the FY '23 outlook. Carrying over from 2022, we see a soft start to 2023 due to market uncertainty and performance challenges at Appen. We -- to help remediate, there is an immediate focus on driving disciplined and greater operational rigor to drive sales and profitable growth. In my first 2 months, I shared with you earlier a dedicated slide on what operational rigor is going to look like, and we're going to continue to work towards that. We have identified annualized cost savings of about $10 million. However, the benefit from the cost savings and increased operational rigor is only expected to commence in second half of '23 and continued growth in 2024. We will continue to invest in product development to the level of approximately 10% of our FY '23 revenue.
To me, if we're going to be a fast growth S-curve organization, products and Engineering is very important to us. Our FY '23 EBITDA outlook is under review and first half of FY '23, underlying EBITDA is expected to be materially lower than our first half 2022. We are also withdrawing our FY '26 targets, pending full strategy review to be shared with all of you in May. As I said to you, this is my first 2 months at Appen. I'm looking under the covers making sure where our financials stand, where our clients are, and I look forward to spending a lot more time with you to share.
To wrap up on this final slide, we have covered a lot in the short presentation. I look forward to sharing a lot more with you, but I felt that I needed to share a lot of this detail with you because it's important as I mentioned, ChatGPT is getting to 100 million users in the 2 months I've been here. I need to really share a lot more details with you so that you can measure me on my say-do ratio. So when we come to May, we're able to have a lot more in-depth discussion.
If I leave you with 5 takeaways, number one, AI is changing the world. Generative AI is creating new experiences like we have not seen before. The future is with Appen, and it's time for us to execute and match towards that growth. That's number one. Number two, Appen is well positioned to capture growth. Generative AI requires human feedback to turn impressive AI into next-generation customer experiences for our growth. We have been doing a lot of human relevance work already. It's very applicable to generative AI. Number three, ethical and unbiased AI is increasingly important. And as I talked about Appen's AI for good strategy of being good, doing good leading good is aligned to support that shift. And finally, our growth question is to improve beyond data for AI into an industry verticalized AI solution and products, expanding our TAM by 20x and that allows us to really give and provide and establishing an operational rigor is a key priority to improve our revenue visibility and funding our growth.
Thank you again for listening. Now I would like to turn the call back over to Melanie, our operator, to open it up for Q&A, please.
[Operator Instructions] Your first question comes from Josh Kannourakis with Barrenjoey.
My first question is just around the cost savings. So obviously, within the New Market segment, still losing a significant amount of cash. Can we just talk about the -- I guess, firstly, the longer-term sort of economics of that business and whether you're comfortable on getting a return for that level of investment? And then secondly, just in terms of the timing of those cost savings just when the full hit of the $10 million will be seen through the accounts?
Josh, nice to meet you. Look forward to meeting you in person at Barrenjoey's. Let me maybe start and I'll turn it over to Kevin, if that's okay. So number one, Josh, I would say for cost savings in the New Markets, I'm not pleased with how our New Markets have grown. Our New Markets is made up of Enterprise, China and our -- some of our Global business. I will tell you, China, I'm pleased with. They've had a great growth, and they're on a good trajectory for this coming year as well. Enterprise, not so much. We need a lot of improvements there. And I would tell you the Global is the big problem that we saw this last year and because some of the areas that we were focused on. But those are -- the New Markets is where the growth lies for us, and we have to make sure that we're driving efficiency in those areas. There's lots of potential in the Enterprise, as I said. And the operational rigor that I talked about will be super helpful for us to drive Enterprise back to growth. I am in China in a month or so to meet our China team as well. And what I'm seeing is very positive signs there.
Kevin, I'll turn it over to you to talk about the cost savings and the timing of the cost savings.
Yes, sure. Thanks. Josh, so yes, definitely, the challenges that we've experienced with this division is operating scale and how we get scalable delivery solution. So as Armughan talked about, we are looking into and working towards target operating models to make sure that we can get that operating scale through the business. At the same time, while we're driving headline sales and sales activity. But the costs we've identified are twofold. The one is obviously to allow us to prioritize and invest in future areas. And the other one is, obviously, as part of rationalizing the delivery model and getting us closer to that optimized structure.
So as we've called out, we expect to start seeing some of that benefit flow through into the second half of this year. And a lot of that is coming from automating some of the delivery capability and function, along with also seeing how we can maximize the offshoring capacity in order to do that. So flowing through starting H2 and then continued improvement through to '24. In addition to that, as Armughan talked about, through the rigor process, the zero-based budgeting and the cost focus, we are going to be looking for additional opportunities to see how we can improve the operational performance and the bottom-line contribution from the New Markets business. I think the expectation, though, is that this is a medium-term play there. Obviously, we're looking to drive growth in sales, looking to get the right structure, right? It's going to take us some time to get there but we do see this being a meaningful sector for us and making a meaningful contribution to us over time.
Got it. So just to clarify on that. So in terms of the $10 million, that's the sort of initial focus with most of that, hopefully, to be seen across the FY '24 calendar year. And then there's also, as we'll get further information for your additional potential savings, but haven't been identified or efficiencies that have been identified as yet?
Yes.
Okay. Cool. Just on that while I've got you, Kevin. So I guess, just you mentioned the debt facility as well, like obviously, this first half has -- the end market conditions continue to be challenged, which, I guess, would put you in sort of free cash flow negative position. How are you sort of thinking about that in terms of usage of debt in the first half, whether it will be required? And also just in terms of the broader, I guess, in terms of comfort into the second half, like maybe you can -- you or Armughan could just give us a little bit more comfort in what you're seeing for clients. You mentioned the second largest clients improving, like just maybe some underlying trends that gives you a bit of confidence that the second half may be better?
Yes. Okay. So let me get into the cash flow stuff first and then we can talk about some of the other things. But as an overall -- I mean, there's a couple of components to this that I think is useful to share. The first is we have facilities that we were paying for but not using. And so in terms of when you think about '23, what we've done is we look to align the facilities to our current needs. At the same time though, we will be conducting a full refi process in the second half of '23. And that's when we obviously have a better view as Armughan refines the strategic view to align with the vision that he shared today. We're going to look to that -- to look that refi process to make sure we've got an appropriate debt structure to support the current and the future needs of the business.
When we think about how we look for this year, what we've done is we've modeled out cash flows along with sensitivity scenarios. And these models show that we have sufficient cash levels to fund the business. We do have interim month working capital requirements, and that may require the use of working capital facilities from time to time. So Josh, that's how we're thinking -- that's how we're seeing the debt structuring for this year. That's how we're seeing that supporting the business.
And then in terms of what we're seeing so far, yes, we -- that call out on our second largest customer was definitely a highlight for us. It's definitely a step change from what we've seen in the past. But we have called out, obviously, that we're seeing the slow start to '23. And the 2 key drivers for that is it's a carryover from what we saw from a macro point of view, but also acknowledging that internally from a performance point of view, there's room for improvement there.
Your next question comes from Garry Sherriff with RBC.
Armughan and Kevin, three questions. Firstly, the outlook on the sales run rate. I mean you guys were up 11% on -- in Q4 on Q3. I guess the question is you've seen January and February trading and considering the pipeline and your deal conversion expectations, do you think the market should expect that similar level of growth for the full year? Or do you think that's going to be better or worse? Just trying to get your initial thoughts from a run rate perspective from Q4 through the rest of the year?
Yes. Garry, so obviously, historically, Q4 is a strong quarter, and it was up on Q3 as expected. I mean, we did have -- it was a tough year last year. We did have order book and POs, but we did have recalibration from our major customers in terms of how they directed their spend. So in the backdrop of all of that, the fact that we had the Q4 uplift on Q3, I think, was a pleasing result. I think when you think about how to interpret that going forward, where we are at the moment is that some of our major customers are still finalizing budgets and stabilizing their cost basis. So as a result, and you've seen we haven't provided order book numbers because of the back of that, the order book in terms of what we're seeing at the moment, we provide the less accurate representation of expected performance than it did a few years ago.
So we're definitely seeing some of our major customers have not yet finalized their spend levels for this year. And so as a result of that, it's -- we are certainly running behind in terms of that visibility. So it's probably a bit early for us to give you any sense. But I think the patterns of what we saw in terms of Q4 over Q3, where we had better revenue and lower costs, I think that's a good starting point and obviously provides a good backdrop for what we're looking to do this in terms of driving better performance, focus on cost and focus and service activity.
Garry. It's Armughan. Yes. Nice to meet you virtually. Are you Sydney-based? I'm hoping that I'll get a chance to meet you as well.
Yes, yes. We will see you shortly, look forward to it.
Okay. That's great. And I'll just maybe just provide some color on to what Kevin just said. For me, I was in here for Q4. And -- but what I've seen in Q1, it requires a lot of rigor, that operational rigor of this week, this month, this quarter. We used to measure this business on a monthly basis. We're going to start measuring this business on a weekly basis. So I'm a week 13 type of person. So week 1 to week 13, people have to call their commit on weekly basis, and they have to deliver their commit. And linearity is a huge, huge aspect of our focus. So in order for us to drive predictability in our sales and what you were just asking, it's going to require that type of rhythm. And -- so again, just getting my sea legs under me. And in May, we'll give you a lot more data on how -- what our trajectory looks like on that, okay?
Yes, yes. No trouble. Last 2 questions. Competition pricing and then leadership for your business units. So looking at competition and pricing, there's been a number of new entrants over the last 18 months or so, and they're competing very hard on price. How do you see you like-for-like pricing of services over the next 12 months? That's the first question.
Yes. So let me maybe start. I will tell you from a competition perspective, we actually don't see a lot of competition in the Globals because we've got some competitors, as you know, they're doing some work. Everyone seems to be in their own silos is what I've found here and long-standing relationships that they have with some of the big techs, and I also think that automation that I just talked about on how you saw that we announced our large language model labeling automation types of solutions. To me, that's going to drive pricing competition in the Enterprise, and it's a big focus for us because driving automation internally would really help us drive profitability and margin improvements. And we have long-term pricing agreements in place with our major customers as well, major clients already, right? So to me, those things are important.
Your next question comes from Ross Barrows with Wilsons Advisory.
I look forward catching up with you, Armughan. I have 2 questions. First one for Kevin, if I may. I guess with respect to the customer concentration, it does remain high at 82%, albeit lower than last year at 87%. And it was mentioned that the second biggest customer grew revenue around 20%. I'm not sure you can really tell us what percentage of revenue each of those clients would represent. But can you say how big the second client is relative to the first just to help us get an understanding of how meaningful that clients increase spend is?
Yes. Maybe I'll give you high-level numbers. Ross, just -- first of all, it's nice to meet you. I would tell you, our first and second customers are -- they're not that far apart, but overall, I think there is an opportunity for us to drive a lot more efficiency in all our top 5 customers is what I would tell you.
Yes. And Ross, for me, obviously, we don't break that out. We have seen, obviously, the concentration come down a bit. We -- you can get a sense from what we've talked in the narrative. You can see there's a bit of 2 speed going on here in terms of we're seeing something from one particular customer, and we're seeing something from another customer. And you can kind of join some of those dots. But we certainly don't break that out, and you know that, but I think we've given something around how you can determine kind of things to what we're seeing from the one customer versus the other customers.
Yes, that's helpful. Just the second question is around, I guess, the AI market. And I guess the change in acceleration in AI is nonlinear as you've mentioned. Given you have the dual focus of, I guess, operational improvement and kind of turnaround objectives as well as keeping pace with, I guess, the speed of changes in AI and generative AI. And you shared your thoughts a little bit, but maybe can you expand on that just in terms of how you can see Appen's ability to retain, I guess, some of its leading positions or even become a leader in some other opportunities that are being flagged and balancing that with the organizational turnaround in certain areas, but other companies that have less of a legacy issue, how quickly they'd be moving. So just that competitive advantage or the ability to create that would be great.
Yes. It's Armughan here, Ross. So I would tell you, I think a few items. One is the Enterprise piece that I told you where I met with, I don't know, almost 2 dozen CEOs, Chief Information Officers, Chief Digital and Chief Data Officers across U.S., Europe and Asia, and many of them are essentially just starting off on their AI models. So it's almost reminds me of where the cloud was many years ago, where a lot of Enterprises were like, yes, should we get on the cloud bandwagon or not. And the big tech clients of ours had already gone to the cloud themselves and then they started offering cloud services, right? That's when many of the companies started to move to the cloud. And the reality is many of the large Enterprises still have not moved all their workloads to the cloud.
So why am I telling you about the cloud? I'm telling you this because AI is in a similar journey where when I go in and talk to many of these C-suite clients, they're telling us that almost 80-plus percent of their data is not structured, right? That is a huge difference than the big tech clients. They almost have 99% of their data that's structured and they're driving insights out of it. So one, I see a huge opportunity for us to offer this AI platform services that I'm talking about today with you, the house slide that I sort of went into. So that's one. So I see that as a huge TAM for us and the opportunity to grow, but that requires a consultative selling approach, making sure that, that rigor that we have with our Enterprise teams that they're moving upmarket there.
Second, I would tell you, we're using generative AI in our internal operations to improve the profitability and the margin profiles of the business. There's lots of generative AI that requires similar services to what we provide. And then every company is going to need to change because of generative AI and the impact will be profound. A lot of the support is needed in this transition for companies move towards. We can't just buy off-the-shelf software that they have been buying to help them automate like ERP or CRM or workflow automation software, if you start reading about what people are predicting and what's going to happen to the software or SaaS industry, it's a very good indication that using generative AI, there's going to be new DIY types of services.
There's a reason that many of the banks still take 6, 7 weeks to do a KYC on a commercial client. Generative AI can cut that down to 4, 5 days. But in order for them to do that, think about the cost that they spend on it, right? So I'm just using it as an example, I can give you consumer retail examples and coming from KPMG and building out a lot of the digital and data for many of the top clients in the world. And then before that, building high-performance computing at Dell and their cloud solutions, I've seen this movie before. And I think these services don't exist today, and we're moving there. And that's why we see the opportunity is immense for us. Hopefully, that answers your question, Ross.
Also Melanie, I think Garry Sherriff also had other questions that we didn't allow him to ask. So we're happy to answer any questions for Garry, because I think he had a few more questions there.
Your next question comes from Garry Sherriff with RBC.
That's great.
Yes, I just -- sorry, I just wanted to clarify, you mentioned that you had long-term pricing lists for your Globals. I just wanted to confirm that they haven't changed because I know there had been previously, certainly pricing pressure from your Global. So I just wanted to confirm that your long-term pricing lists haven't changed?
Yes. And Garry, that comments in reference to and I think we did announce this to the market a few years ago. That was a multiyear price agreement with a large customer of ours. So it's not -- we have rate cards across all the Global customers. But with this particular customer, we actually have fixed price agreement.
Okay. So for that one, yes, locked in, but others perhaps not by the sounds of things?
Yes. Look, they move up and down. I can say this, which I think can help a little bit, and that is -- and because it also comes up with a question in terms of how do we manage cost inflation and how do we manage this? Well, with this other -- once again, once of our larger customers, we've actually recently had a rate increase in order to address inflation issues and to make sure that we can continue to provide a good quality and a solid supply of services to this customer. So we actually have actually had a price increase from that one that customer. Yes.
Got you. I guess, all right. If I take a step back then, on average, do you think the price has gone up or gone down across your Globals, maybe that's a simple way of doing it?
Well, I mean, as we said, with one of them, it's fix, there's no movement. The other one had an increase and the other ones, there's a bit of movement today. I think we've talked about it before. There's obviously every couple of years, there's an auction process or a tender process or whatever it is and we reset pricing in the normal course there.
Okay. And just the last one, Armughan. Given -- I mean, again, you haven't been here for long, but talking about leadership for your business units, do you think your leadership team at present is fit, I guess, the initiatives that you're going to embark on? Or should we potentially expect changes in your business unit leaders over calendar year '23?
So Garry, as I said, this is what week 9 for me on the job, and I have been looking at seeing what we have. And I would tell you, we have some talented people in this organization, a good foundation, and I'm in early days, and I'm working with the teams and assessing our talent. I'm here to look to reset this business and drive this operational rigor, and we have the right people in the right roles. It's something that determining and we're going to work towards getting to growth. Numbers, revenue and profitability speaks louder than what we have, right? So that's the most important thing here.
[Operator Instructions] Your next question comes from Siraj Ahmed with Citi.
Kevin, just 2 questions. The first one, just in terms of the outlook for FY '23. Given Armughan -- given your flagging changes in operational rigor, should we think that you get back to sort of top line growth this year? Or is it a bit too early?
Yes. Siraj, it's nice to meet you, Armughan here. I would tell you, I think, as I'm assessing the business, the reason I am resetting our FY '26 targets and looking at what the first half and the indication we just gave the first half, to me, I need time to assess this business because, again, it's very lumpy, as you guys have seen over the last 1.5 years of us missing targets. We need to really better understand.
I give the analogy of like I'm flying this plane and I've got a bit of blinders on. So I need to start seeing weekly progress on bookings to see how we're tracking to it because we usually look at our revenue on a monthly basis. I need to see bookings on a weekly basis to drive predictability in the business. To me, that's very important. I think we have a strong foundation, as I mentioned, but it means that operational rigor and -- but there's big, big potential across the business.
I'm telling you it's still early days for me, but all the clients have gone to see, all of them want to work with us. And these deals -- because AI is so new from an Enterprise life cycle perspective, I'm trying to figure out is our timeline for deals is 3 months? Is it 6 months? Is it a year, right? We have a lot of pipeline. Pipeline is growing in the right direction, but we need to start seeing closure rates and because Enterprise is the newer side of the business for us, I need to see predictability and linearity in it.
Got it. Second one, just on generative AI in the potential. I understand -- just in terms of deep learning and what you've done on the relevant side and with generative AI, do you think the need for human input is the same or more or less in terms when you compare that with deep learning?
Yes. Good question, Siraj. So this is a question that I'm getting a lot these days. And let me try to explain. That's why I added that slide in there to try to get people to understand that the difference between deep learning and generative AI, first of all, before AI, most of the clients who are working on data and analytics, and they were just trying to get all their data into one place, but then they started working on deep learning and in deep learning for a lot of clients to get their data ready for AI. They had to do the data collection, the data preparation, which is labeling and then run it through the model and then do the data relevance work. So many of the clients are still working in deep learning mode, right?
And so we -- interestingly, the 3 areas I just told you that we work in almost 70-plus percent of our revenue comes from the data relevance work, which is once they have run it through the MLOps models, they then do relevant checks on it. In generative AI, that relevance check still matters in a huge way, right? So need for human input is critical to get the large language models working well. You have seen what's happened in the last 4 weeks of ChatGPT, hallucinating and many of the other models hallucinating and that hallucination is becoming a huge problem with generative AI.
And if you've seen search and ads were very limited to tech, and they were very concentrated. But when you look at generative AI, it has much broader applicability. And the potential is large but imagine if you're asking ChatGPT to make your financial decisions of where you should put your children's money or your retirement money and it's giving you hallucinated answers, those are cute answers. They're not game-changing answers, right? So that requires a lot of relevance checks for Enterprises and even these large, big tech companies are realizing that they need to have the same relevance check. So net-net, what I'm trying to tell you is that potential is very large but unclear how it will compare versus our current relevance business. But today, that's -- we have seen many of our clients engage us already in relevance work in generative AI.
Your next question comes from [John Hibbett] with Hibbett Superannuation Fund.
Thanks very much for the presentation. I appreciate that. I have 2 questions, which are complementary, so I’ll ask them both together, in order for you to give a more comprehensive response. The U.S. government has been applying restrictions to technology operation with China. What’s the risk that they will apply this to Appen? And if so, how will you respond? And the reciprocal question is the Chinese government has been stepping up – stepping into increased influence in successful companies. So what’s the likelihood of that occurring to Appen? And if so, how will Appen respond, please?
Yes. So John, nice to meet you. Armughan here, and I’ll turn it over to Kevin to maybe add some color to this. I will tell you that our China business is growing amazingly for us. We are the number 1 market share leader. If you look at the number 2 market share leader, Speechocean and they’re publicly traded company, you can see where their valuation is trading versus we are globally. I would tell you, I see our business continue to grow in China. I’ve worked in my previous slides, if you looked at my background, between 3Com, and we had a joint venture with Huawei-3Com.
I’ve seen a lot of this movie before with CFIUS in the U.S. And right now, the structure that we have put together is a balanced structure. And to me, I’m not worried about any dependency on where that is. We have separated our business in China, including our tech stack to minimize the risk. As I said, China is an independent entity, and we are continuing to monitor our risk very closely. Kevin, do you want to add anything?
Yes. No, I don’t think there’s much more to add other than we – from when we set up China and when we set up our federal business from day 1, we’ve been extremely purposeful in terms of that setup knowing what the requirements and regulations and concerns of the U.S. government is, together with what are the concerns from the Chinese government. So it’s something we’ve been across from day 1, and we’ve addressed that accordingly. Thanks for the question’ John.
Your next question comes from Wei Sim with Macquarie.
I've just got one question, and that's just regarding our views on project versus committed revenues going forward. I think we did have the view previously that we wouldn't focus as much on committed revenues, and it would be more so on the total amount of revenue growth rather than committed. Armughan, I'd be keen to hear your views as to the quality of the revenues, whether that's something that matters to you or it's just about the absolute growth?
Yes. I'll just give some key points and then hand over to Armughan because I think it's central to everything we've been saying about in terms of driving top line growth and sales rigor. But I think based on the bottom line for us is we've got one objective here. It's clear. It's driving top line revenue growth. The committed -- we had some inertia, some challenges before when we were dogmatic around committed revenue, and that hurt us. So essentially, our focus is on revenue growth, first and foremost. And we're less concerned in terms of how it comes about. But just hand it to Armughan because as I said, I think it's central to a lot of his focus areas and driving top line growth.
Great. Thanks, Kevin. Wei Sim, nice to meet you virtually. Look forward to meeting you as well. I would tell you a few things. One is, if I look at our relationships and the clients that I've gone to meet so far, we have 9-year plus relationships with our clients. To me, that gives me a lot of surety on how our clients value us. Many of the projects are ongoing for many years. So this is not like they wake up every day and they decide to give us business or not. And then our deep relationships with our customers is a huge strength to Appen. As I mentioned earlier, the operational rigor, driving top line growth and a profitable growth is really going to be the focus area for me.
Okay. Great. So maybe Melanie will end the questions here, and I can wrap up quickly.
Please go ahead.
Thank you, Melanie. So first of all, thank you for all your questions. Sorry, we ran a bit over, but this is my first one. So I hope you will pardon me for that. But I really wanted to address all of the questions that all of you had. I'm really looking forward to meeting all of you, Josh, and Garry and Ross and Siraj and John and Wei Sim. Looking forward to meeting all of you to build relationships with you and get a chance to explain any other questions that you have. I think we have some one-on-one established. We also have our May Tech Day coming where we'll get a chance to go into a lot more of what I call the say-do ratio, what I told you today and what we're going to do.
I just want to end with this. Number one, AI is changing the world and generative AI is creating new experiences that we have not seen before and future is what Appen, and it's time for us to execute and match that growth. And Appen is well positioned. Our leadership team is well positioned. Our employees that we call Appenites and our crowd team are super excited.
I've had a chance to pretty much touch and -- most of our 1,500-plus employees, along with many of our crowd members. And I'll tell you the kind of talent we have here, it's super exciting to me, and it enables me every day to come to work and can travel nonstop and make sure that we are doing good work for them to make them proud. And I know we are very well positioned to capture this growth. We have gone very quickly from tens of millions or hundreds of millions. And as we go from hundreds of millions to our new growth targets, we really need to move towards leveraging the ethical and unbiased AI that I talked about AI for good. Do a be good company, to do good company, to a lead good company. And then finally, to achieve our growth vision that's going to require a lot of our rigor that I talked about, and I look forward to sharing a lot more with you. So thank you for your confidence in Appen and your time today. This will conclude the call. Thank you so much.
That does conclude our conference for today. Thank you for participating. You may now disconnect.