Good morning, ladies and gentlemen. My name is Chris Fondweller and it's my pleasure as Chairman to welcome you to this virtual Annual General Meeting of Appen Limited. Due to the current COVID-nineteen pandemic, we thought it prudent to take steps to discourage a physical public gathering and encourage participation online. We hope that holding a virtual meeting will assist in further curbing the spread of COVID-nineteen virus and encourage greater participation and engagement with our shareholders. So thank you for your cooperation and patience in these unusual circumstances.
It's now just past 10AM, the nominated time for the meeting, and I've been informed that a quorum is present. I note that the meeting has been validly constituted, and I'm pleased to declare the meeting open. There are three components to today's meeting. First, I'll provide you with an update on the business from a strategic perspective. This will be followed by a detailed overview of the group's performance for the 2019 financial year by our CEO, Mark Brayan, and an address from the Chairman of the Nomination and Remuneration Committee, Bill Pulver.
We'll allow questions time for questions on the business at this time. Shareholders who wish to ask a question, please click on the Ask Question button, type your question and press Submit. I encourage shareholders attending online who have questions to send their questions through as soon as possible. We've received a number of questions prior to the meeting and these have been largely addressed during the upcoming presentations. Following the general business questions, we'll progress to the formal business of the meeting, where the resolutions provided in the notice of meeting will be put to shareholders.
And we'll allow time for questions and answers regarding these specific resolutions before proceeding to vote on them. So I'd like to begin by introducing my fellow directors that are present online today. Bill Pullman, who's a nonexecutive director and Chair of the Nomination and Remuneration Committee Robin Lowe, Non Executive Director and Chair of the Audit and Risk Management Committee Dina Schiff, a Non Executive Director Steve Hasker, Non Executive Director Vanessa Liu, our recently appointed Non Executive Director and of course, our Chief Executive Officer and Managing Director, Mark Brayan. We also have various Appen executives online, in particular, Chief Financial Officer, Kevin Levine representatives from the company's auditor, KPMG the audit partner, Tony Nymak and representatives from the company's share register, Leak Market Services, who are providing the virtual meeting platform today. So I'll turn now to the Chairman's address.
Let me again welcome you to this meeting. As is plainly evident, the COVID-nineteen pandemic is having a deep impact on many businesses. The effect of the pandemic has been the subject of several questions submitted to this AGM. In Appen's case, as advised in our ASX announcement on April 15, our distributed working model is allowing us to adapt well to this challenge. This model has proved to be a real strength to the company, and Mark will touch on this further in his presentation.
I should note, however, that before we entered this year, we entered this year with a very strong balance sheet and over $100,000,000 in cash resources and that provided an excellent foundation to handle the pandemic. So turning now to the review of 2019, I'm pleased to report that we completed another year of profitable growth and strategic development for our company. Appen has maintained and strengthened its position in the rapidly developing and exciting field of machine learning and artificial intelligence. So the financial results of 2019 were strong. This performance reflects the benefits of scale, technology investment and productivity improvements, which we have driven which have driven our processes.
The 2019 results incorporate, of course, the Figure eight acquisition, which was completed early in the year. So here's a summary of the results financial year 2019. Revenue grew to $536,000,000 an increase of 47%. Underlying EBITDA was $101,000,000 up 42%. Statutory EBITDA was $87,900,000 which is up 29% and that reflects the impact of the transaction costs relating to the figure eight transaction.
We maintained our strong profitability on sales with an underlying margin in 2019 of 18.8%, a slight decrease on 19.6% in the previous year due to the impact of the figure eight losses. The underlying margin without figure eight was 21.5%. The underlying net profit after tax was $64,700,000 up 32%. So the Board declared a final dividend of $05 per share. And together with the interim dividend of $04 per share paid last September, the total dividend payment for 2019 was $0.9 So a few words on the strategy and the outlook.
As you know, we operate in a dynamic industry which presents massive opportunities and of course some challenges. We're investing to capitalize on the unique position which Appen has built in this most exciting sunrise industry. Appen's strategy continues to be focused on the priorities outlined in previous reports to the shareholders. This was further demonstrated in the Investor Technology Day, which with a presentation earlier this month. Our technology development has provided a truly integrated platform, supporting crowd management, client workspace and annotation tools.
The scale of our workflow and the breadth of our software tools are arguably world leading. At the same time, we've sought to broaden our revenue base by investing in sales and marketing to win new customers and expand into new geographies. We're determined to invest and maintain our leadership position in one of the fastest growing fields of technology. In early twenty nineteen, as I'm sure you're aware, we acquired the San Francisco based company, Figure8. Together with our acquisition of Leapforce in 2018, this contributed efficiency and revenue diversity and strengthened our technology resources.
So there are reasons to be positive about the future. The AI market is growing at 28%. Training data, which is our core business, is at the core of building AI models. The deep learning techniques now widely used in developing applications require very large volumes of training data. Fundamentally, the more data, the better the outcome.
Further, to maintain currency, data typically needs to be refreshed, thereby creating further demand. One of our real differentiators from the point of view of our customers is our ability to support all active modalities, search relevance, speech and language processing in over 180 languages and LiDAR image and video computer vision. So our priority is to accelerate growth in new industry sectors beyond the traditional tech sector customers, for example, government and in new geographies. I'll turn now to environmental and social responsibility. Appen is a truly global company and we celebrate our diversity, as do our customers who value this as a core differentiator.
Our work from home model provides a source of income to individuals that may otherwise find this difficult including the disabled. During 2019, we launched an initiative to develop future leaders to improve gender balance in our senior ranks. The Board is committed to this initiative and to diversity in general. An important emerging characteristic of our industry is the societal impact of artificial intelligence, including ethical implications and privacy issues. We launched our CROWD code of ethics in 2019 to ensure fair pay and treatment for our at home workers.
We're also working with our customers to understand and address ethical issues around the development and deployment of AI, including the need for unbiased data sets. The nature of our knowledge based business results in a relatively low environmental footprint. The work from home model for our crowd workers reduces carbon impact and we seek to use video and phone conferencing to minimize travel, of course, more so than right now. We bought carbon credits to offset the impact of our air travel footprint in 2020. Management and employees.
The exciting challenges of the AI machine learning industry place enormous pressures on app and management. Our customers are sophisticated and demanding and the pace of change is intense. Our management team under Mark Brand's leadership continues to meet these challenges daily and maintain Appen's position at the front of the growth curve. At the end of twenty nineteen, we had seven eighty one employees, which was up 52% from the previous year and is still growing. On behalf of the Board, I thank our management group and our staff for their contributions.
Our global workforce of more than 1,000,000 skilled contractors is a key asset for Appen. They live in 130 countries and they speak over 180 languages with cultural and ethnic diversity. They underpin our success at no time more evident in this difficult COVID-nineteen pandemic. As a truly global company, we need to compete for talent in a highly competitive industry sector. We seek to remunerate managers fairly and competitively with short term and long term incentive schemes aligned with long term shareholder wealth.
And Bill Pulver, the Chair of the Nomination and Remuneration Committee will expand on this later. The Appen Board has a busy agenda and is engaged in all aspects of Appen's business. We need to ensure that the Board has the optimal mix of experience and skills for this fast moving industry. So we're pleased that Vanessa Liu accepted our invitation to join the Board in March. Vanessa is based in New York and brings highly relevant experience in U.
S. Markets and global technology sector. And we'll hear from Vanessa later in this meeting. So I take this opportunity to thank my fellow directors for their contribution and I'm grateful commitment. Finally, thank you for your loyalty and support as shareholders.
We value this highly, and we give you assurance of our continuing efforts to make this company even more successful. So ladies and gentlemen, I'll now hand over to Mark Brayan for his CEO presentation.
Hello, everybody. I'll echo Chris' welcome to the meeting, albeit a different format of meeting from our usual in person meeting. And I hope also that everybody is safe and well during these most unusual times. I've posted the presentation to the ASX, so you can refer to that as we go along. And I'll guide you through that over the next little while.
So to the first slide, we operate in the artificial intelligence industry and AI is a very important industry in terms of the development of current and future products, but it's also the industry that companies are turning to improve their productivity and efficiency. And we expect this to accelerate post pandemic as companies digitize more aggressively to improve their businesses and recover after the pandemic. Artificial intelligence relies on data to learn. It learns from data. It is trained by data.
And our role is to provide the pivotal data that makes AI work. Arguably, as the world's leader in data provision for artificial intelligence, we could plan to make AI work in the real world. We're in an exciting market. As Chris said, the market is growing at 28% per annum, and we operate in all data modalities. We provide audio speech data, we provide image data, video data, now LiDAR, which is very important in the development of autonomous vehicles, as well as text data and search relevance data.
And this has worked well for us, us as a company, our customers, our staff, our crowd and of course you, our shareholders, you can see that our growth both in revenue and earnings has risen steeply over the last five years as we've been in the public market. So we're selling into a market and we're executing reasonably well. In terms of the future, we're very well positioned to win. Our 1,000,000 strong at home crowd is ideally placed during the pandemic. We have very broad, deep expertise, and we have our technology suite of which more later in the presentation, which is fundamental to the growth and scalability of our business.
So over to Page five, and Chris has taken you through the results, but I'll just touch on a couple of highlights. First of all, revenue growth and underlying EBITDA up 4742% respectively on last year. This is tremendous growth, particularly as the company gets bigger and the numbers get bigger that we're maintaining these high levels of growth is especially pleasing. Yes, there was some impact on the margins year on year, but as Chris explained, that was due to the acquisition of Figure eight, which was a highly strategic, very important acquisition for us. On the right hand side of the page, we point out the organic growth, which is illustrates that the underlying business continues to perform very, very well.
So whilst Figure eight added some revenue to us, revenue organically was still 37% and underlying EBITDA growth was 51%. And of course, with EBITDA outpacing revenue means our margins are rising and underlying EBITDA margins were 21.5%. So not only is the business growing, but we're improving the business as we go, and we'll talk a little bit more about that when we get on to technology. Each of our divisions is performing well and delivered good growth on last year. So speech and image, which we used to call language resources, provides high quality speech, natural language and image data for our clients, and that's growing very well year on year, up 60 sorry, up 32% on a revenue basis.
And the majority of that revenue came from our existing customers, but of course, we're landing new customers on a regular basis, even more so now with the Figure eight platform, which opens the market for us and enables to serve even more customers and opportunities. Search relevance grew handsomely at 37% on a revenue basis and is up now to $430,000,000 per annum. The reason that the relevance data outpaces the speech and image data is that relevance data ages very quickly. So we have to replace that data over and over consequently, our customers have almost an ongoing need for that data, and that continues to rise. And that data powers the world's biggest search engines, social media engines and ad placement engines.
It's fundamental to the success of those companies. And we'll talk a little bit more about that when we get on to the impact that the pandemic is having on our business later in the presentation. To figure eight and very pleasingly after the acquisition at the beginning of at the end of Q1, sorry, last year and a flat Q2, we saw good growth in Q3 and Q4 as that business bounced back through lower churn and more new customer wins. So we're very excited about that, and it's now a very fundamental part of the business. As of January, when we completed the earn out, we started the integration of the businesses and I'm pleased to report that that integration is going very well and is almost complete.
So the final earn out payment of $39,000,000 led to a total acquisition cost of $286,500,000 which is 5.7x 2019 revenue. And for those that are familiar with the valuations in The U. S. On technology businesses, that's a fairly low revenue multiple. Back to the integration, we've now unified all of our teams, our sales teams, our engineering teams, our client service and customer success teams are all unified under single leadership and the teams are working well together.
We also are going to market with a new product offering. We've refreshed our look and feel, unified our branding and integrated our back office infrastructure. So the integration of that business is going very well, and we're very pleased with the contribution that the Figure eight team and the platform and their customers are making to our business overall. We expect that integration to be fully complete by the end of this year. Turning now to our cohort chart, which you've seen before.
This chart illustrates the revenue that any given cohort of customers provides in one year and in subsequent years. And so you can see that the, particularly those, reddish colored bars, as they progress from year to year and grow from year to year, that indicates that the customers that we derive revenue from one year continue to provide revenue for us year on year, which goes to the importance of the revenue that we provide and also goes to the repeatability of the revenue that we derive from our customers. So you can also see the impact of the figure eight revenue in 2019 and how that contributes to the overall group. To non financial outcomes, and Chris touched on these as well, our environmental footprint is fairly low for a business of our size. We are offsetting the travel that we take through carbon offsets.
The societal impact of our business is quite impactful, mostly because we have a very large crowd of workers that contribute to the provision of the data that we provide for our customers. We committed through our crowd code of ethics to pay our workers fairly, and we're very pleased and very proud to the fact that we are paying our crowd workers fairly. We are also a member of the Global Impact Sourcing Coalition. Now the coalition provides work opportunities for people largely in developing countries that may be disadvantaged in some ways. Our contribution to that is a growing and I must say very enthusiastic, team in The Philippines who have hearing disabilities.
And they just as an aside, when we took all of our Filipino workers to work from home, there was a little more work required to get them set up properly and logistically with bandwidth and computers and so on. But our hearing impaired team created a lot of in house videos that were all done with signing. So everybody knew what was happening and what they had to do and how they could work more effectively from home, etcetera. Was very heartwarming to see the effort they put into that and to see the enthusiasm in doing something as simple as moving from the office to the at home environment. So we're very pleased and very proud of that team.
We're also doing some work with an organization called Translators Without Borders. We're working with some of the global tech giants to help them train their speech recognition models and their translation models to work in the languages of developing nations, the project and to include terms that are specific to the pandemic. So the word pandemic, for example, wasn't a word that I used last year. It's a word now on everybody's in everyone's vocabulary. And if the pandemic impacts, as the world expects it will, impacts developing nations.
People are going to need to translate things online, how they care for themselves, etcetera. And so we're working with those companies to help them improve their machine translation models to include terms that are specific to the pandemic for those communities. And we're very proud to be doing that. From a governance perspective, of course, we welcome Vanessa to the business to increase the independence of the Board, and you'll hear from Vanessa later in the meeting. Okay.
On to our growth initiatives, of which there are three that we've talked about in prior meetings. First to the government sector. We view the government important sector for us to contribute to. It is also an area that is investing into AI. And it's also an area that requires specific expertise and specific infrastructure in order to participate in government work.
We've been participating in government work for many years, the provision of speech data to governments around the world, and we believe this to be an opportunity that we can invest into, and we're doing that. We've opened an office in Washington D. C. We've hired David Poirier, who joins us with a very strong military and contractor career in The U. S.
Area. David's on board. He's building the team and building the pipeline. And the full setup of that office is almost complete. We expect that to be fully complete by the end of this half.
And we're very excited about the work that we're doing in that space, and we're very pleased that we can make a contribution to the government space as well as the commercial space. So very pleased with that and investing to plan. Outside of the government, we are also investing in sales and marketing in general and we flagged this in the year when we presented our full year results. So we view the opportunity to sell training data into the AI community as vast and open beyond the large tech companies that we serve. So we now have teams focused on the global tech giants, of course, enterprises in The U.
S, commercial companies in The U. S, certain verticals such as automotive, financial. We are growing our teams in Europe, in Asia Pacific, and of course the government sector as well and China, which we'll get to. So all of those teams are focusing on customers who need training data and we're very pleased to see that the pipeline is building and we're adding customers to the overall customer mix. We view this as the right time to invest in sales and marketing.
Recall last year, we invested heavily in technology. We now have a leading suite of technologies to sell to our clients and so investing into sales and marketing to add to the revenue base and also importantly diversify the customer base. So investing to plan for our sales and marketing growth. To China, very pleased with our progress in China. China, of course, is the second biggest, may soon be the biggest market for artificial intelligence globally.
We have a team on the ground of over 100 people who are supporting customers in that market. Their focus is unsurprisingly the large tech companies, and the difference that we bring to the market is our global footprint and the ability to provide high quality global data into that market. We've set up our operation in China to be quite separate to the rest of our business for reasons of culture. Businesses in China run differently. They run quickly, I can assure you, but also for reasons of IP and data protection.
So our China team has its own data stack sorry, its own tech stack. So Chinese data remains in China. Rest of the world data remains in our rest of the world systems. There are very stringent laws around data moving across borders and we respect those laws and keep things very separate for that reason. But overall, we're pleased with the work that's going on in China.
We're adding new customers to the business. Our team in China is back in the office post pandemic. They did lose a little bit of momentum earlier in the year, but it appears now that things are getting back on track. And it's interesting for us to have that team there, a little bit ahead of the rest of our business in terms of responding to the pandemic and its recovery. So we're pleased that it's starting to reaccelerate post pandemic and very pleased with the work that the team is doing there for us.
Okay. A little bit now on the market opportunity. And Chris mentioned or two important aspects about our business, and that is that artificial intelligence relies not only on large volumes of data, but on ongoing data refresh. And I'll talk through each of these slides because they're quite impactful and quite profound for our business. On the left hand slide shows the performance of the networks, and how those networks improve with increasing data.
So on the vertical scale, we have the performance of the network. On the horizontal scale, we have the amount of data. And you can see also that there are different models, regular machine learning, for want of better expression and then neural networks all the way up to deep neural networks or deep learning is an expression you may have heard. And deep learning, they're the networks that work the best, but they're also particularly data hungry. So you can see their performance exceeds the other networks, and you can also see that the more data they consume, the better they perform.
Now the amount of data they need is vast. The study in the middle, this was a computer vision study, and what it looked at was how accurate could identify a picture. And I'll give you some examples in a second. But they had the original accuracy of the model, with 10,000,000 images to build that model was around 62%. In order to get that model not even to 80%, to 77% accuracy, it required 30x the amount of data, so 300,000,000 data points.
So if you've used, a product, a speech recognizer, for example, and it doesn't recognize your voice and you get a little frustrated with it, it can be improved with more data. But as you can see from this example, it takes an enormous amount of data to improve. Now if I was the developer of that product, I would want it to be better than my competitors. So there's a race between the developers of AI to build the best products. And correspondingly, there's a race to get and use more and more data, which, of course, is very good for our business because as our clients need more data and more specific data for what they call edge cases, a speech recognizer may work fine in, let's pick an East Coast U.
S. Accent, but it may not be so good with a Southern U. S. Accent. So they have to collect data, and we often do provide data in both of those accents, even though it's fundamentally the same, it's definitely the same language, and it's in the same country.
So the more data and the richer that data, the better the model performs. Now also on the right hand side, you can see a chart that illustrates the phenomenon of model decay, the model being the underlying engine that the AI relies upon. Over time, the models can atrophy. They can lose performance. And so the data needs to be refreshed.
So again, for example, the work we're doing with Translators with Borders, we have to update those models with terms that are specific to the pandemic to make them useful for the current time. More impactfully, particularly for our business, is that relevance, which forms the majority of the business we do for our clients, that ages very quickly. And so we are constantly providing relevance data for our customers. So this slide illustrates the market opportunity for us. It illustrates that models need data.
It illustrates that they need enormous amount of data. And it illustrates that some, 34% of models need to be refreshed monthly, some models need a lot of data refresh, all of which is good for our business. Now a complexity of the data and the data types that we work in is we also need to involve human beings to collect the data. So I have some examples here. And these examples show pictures of dragonflies and show what the human thought was in the picture.
In the first picture, and it may be on the top left hand side of the screen if you're reading the presentation from the ASX or it's central to the screen if you're watching it online. The first picture, the human said, well, that's a dragonfly. For some reason, maybe the colors, the artificial intelligence that was applied said, oh, that's a skunk. So that's not labeled correctly or it's not being predicted correctly, sorry, from the AI. The next answer shows a dragonfly on the handle of the shovel and the human said it's a dragonfly.
The AI said it was a banana. So got that completely wrong again because of the color. And the next example, human says it's a dragonfly on the edge of a flower pot, and the AI said it was a sea lion. So I credit the AI with some imagination. It's actually doing quite well because I couldn't have come up with a sea lion if I saw that.
And the last one, human says dragonfly, AI says mitten. So you can't you do need to use humans to prepare and label data because the human cognition is so much more sophisticated than the machine. The machine gets better with more data, as we've explained. But in order for the machine to learn, it has to have this labeled data. And so this is a very simple example of data labeling to picture what's in the picture.
Typically, though, the things that we do are more complex. And I'll show you a couple of examples, and I hope the technology keeps up with it here because we had trouble with this the other day, but please bear with us. So this scene that you should be seeing is an image out the front of the vehicle from a driver's perspective. And this use case that this data is going to be used for is clearly an autonomous vehicle. So for the autonomous vehicle to work, it has to be able to recognize in this image all the elements on the road and around the road in order to drive safely.
Just as we would if we were driving the car, it has to know to stay on the road. It has to know to stay in the lane. It has to know, what the markings on the road indicate. It has to not hit other cars. It has to definitely not hit people.
And it's probably best if it doesn't hit a or a tree. So it needs to know what all of these things are. So how do we help the machine learn? What we do is we identify each feature and we tell the computer what that feature is. And the process looks something like this.
So a user would go to the video, and they would then, one by outline each of the features, in this case, starting with the vehicle. So you can see that it needs to be fairly precise because we don't want the car to know that the sign on the road is the vehicle. We just want the car to know that, that bit covered in blue is the vehicle. And then you go to the next vehicle and the annotator or one of our crowd workers, who is one of our crowd workers, sorry, identifies the outline of that vehicle and then tags it as a vehicle and so on and so forth to other features. So for example, we have to make sure that the machine knows where the footpath is or for the Americans, the sidewalk.
And in this case, instead of drawing a polygon around it, we're coloring that in. Everything that's colored in pink is the sidewalk as far as the computer is concerned. And there are ways that we can correct it because we've sort of overrun down the front there. So, we'll go back and just tie that up. And each feature, the shrubs, the bus, the line, everything needs to be identified.
So it takes a while. So when we start talking about vast tens of millions, hundreds of millions of images and then we look at what's involved in labeling these images, you can see that it's a really labor intensive task. And so the data becomes very expensive. So there are ways that we can improve this. So in this next example, what we've done is we've used machine learning to automatically label this image.
And it's up to the crowd worker to just check that everything is correct. So rather than actually labeling it, all they're doing is checking it. And if I run this demonstration, you can see that the worker now just goes to the and says, yes, that's a car, and then goes to the next thing and say, yes, that's what's the next one they're picking up? Oh, the bus or the truck and so on and so forth around the image. So what these handful of slides and a couple of examples show you is that, first of all, AI needs vast volumes of data to perform well.
And second of all, you need humans to do this, but it's a very, very laborious process. So one of the things that we're doing is we're using a combination of artificial intelligence to pre label the data and then using the human to check that to make our provision of high quality human label data much more efficient and to also add to the scalability of our business. So I hope those small examples, first of all, I hope the technology worked. And second of all, I hope they're illustrative of the work that we do. So I touched on this, but it underpins the importance of our technology.
We have our crowd, which of course does the workforce, but our technology platform is what drives the operations of our business and also drives the scalability of the business. So on the left hand side of the diagram, we have three fundamental components of our platform, the crowd management component, which we do all of the recruiting of the crowd, and we distribute work to the crowd workers. We check the work that they do. We monitor the quality that work and we pay the workers all through that platform. It's highly scalable.
We have over 1,000,000 people in our database, and we're paying 50,000 more than 50,000 people every month to do crowd work for us. We also have the client workspace, and that's the interface that the client uses to upload data. If that's required in the project, if we're not collecting the data, the client may need to upload the data. The client can can run their own projects or we can run them for them, or the client can check on how the project is going and the quality, etcetera, of the project. And then finally, the annotation tools, which are the tools that you saw the example of, they are the tools that the worker uses to annotate the data and they are the tools that we are augmenting with artificial intelligence to improve the productivity of those tasks.
So the things that we're investing into to improve the scalability of the business is adding annotation tools to improve increase the number of data types that we address. So not just speech and image, but video text and different modalities. The more data types we can handle, the more opportunities and the bigger our addressable market. Secondly, the investments improve the productivity of the crowd. So rather than have the crowd worker laboriously place a polygon around each vehicle, for example, they can just check that the polygon that the machine has put there is correct.
It's not always correct. As we saw initially, the AI is not good at picking dragonflies, so it's going to make some mistakes when it's picking cars. So the worker goes in and corrects that and make sure it's 100% correct, which is what the client requires, but it is a lot quicker using the pre labeled data. And then finally, it goes to our own internal efficiency, and we've seen this come through in some our results. If you look at the gross margins of the relevance business over time, they've steadily improved and a lot of that's due to the scalability of the business that we've delivered via technology and process improvement.
So overall, the technical platform is a really important part of our business, and it's helping the business grow, both in breadth of opportunity, but also in scalability. To the pandemic specifically, it's something we're all living through and again leads us to do an online meeting and many other changes. But I'm very pleased to report that the business is weathering these changes very well. So from a sort of an ongoing operations perspective and a safety perspective, we have all of our staff working from home. Now we do have some people working in facilities doing secure work.
We were able to get customer permission to move most of that work to at home, but some of it we can't. So we're working with we've got socially distanced crews working in our facilities to look after those customers. The at home crowd is ideally positioned in this environment, and we have been in touch with the crowd, and we are providing them with information to help them through the pandemic. But overall, both our internal and crowd operations really haven't missed a beat through this phase. So we're really pleased and proud with the work that everybody has done to look after our customers, and our operations are running pretty much without a hitch through the pandemic.
It's also very pleasing to note that our customers are quite resilient. Our customers are the large tech platforms and they really are the heart of everybody's lives at the moment. We're all relying on search. We're relying on social media. We're relying on e commerce.
And the world's biggest players in those markets are our customers. And their strength and resilience flows to us and makes our business stronger and resilient as well. Now we do keep our eye on what impacts could occur in our customer base. There has been reported and modeled slowdowns in online advertising spend, which, of course, is the major revenue driver for some of our customers. We've yet to see that flow through customers, but we keep our eye on it.
Just the economy in general and spending on IT and a few other things is likely to impact some of our smaller customers, so we're keeping our eye on them as well. But overall, the impact thus far, based on the current information is fairly small. And then finally, and Chris pointed this out as well, we have a very strong balance sheet, cash balance in excess of $100,000,000 We've got undrawn working capital facilities. We have healthy cash flow and conversion and overall the business has low capital requirements. We do in our outlook statement talk about capital management.
During the current pandemic, we're being very conservative with capital, as I hope you understand. The review of capital is ongoing, but at the moment we need or we would like to keep as much cash as we can for obvious reasons to keep the business very strong. So finally, to the outlook statement and a couple of other important things to leave you with. First of all, despite the pandemic, we view, our business is very strong, and we view the future opportunity as unchanged. In fact, there is potential more opportunity for us post pandemic as people accelerate the use of AI.
So we're continuing to invest as planned through the pandemic. Those investments, as we've explained, will soften our first half margins on historic levels down to the mid teens, but we're expecting those margins for the full year to recover to the high teens. As I explained, we're seeing negligible impact from the pandemic, but of course, we monitor that very, very closely. It's a dynamic situation to all of us, thus far on what we see negligible impact. We reaffirm our full year guidance.
So our year to date revenue plus orders in hand for delivery in 2020 now stands at $350,000,000 and that's as of May. So the order book stands at $350,000,000 as of May. And our full year underlying EBITDA for the year ending 12/31/2020 for this year is expected to be in the range $125,000,000 to $130,000,000 And as is our practice, the forward parts of that forecast from May to December are at $0.70 in the dollar. So that was the rate we struck at the beginning of the year and we maintain that rate for the forward parts of the year. So guidance reaffirmed, which in the current circumstance, we think illustrates very strongly the strength of our business, and very strongly shows how well we're operating and how well our customers are standing up through the pandemic.
And I mentioned with capital management, we're just taking very conservative approach to that at the moment, given the pandemic and having cash in the bank is very, very helpful. And of course, the outlook is susceptible to some upside and downside flows as it says on the slide there. But overall, we're very, very pleased to be able to reaffirm the guidance that we set at the beginning of the year, which we think is a real illustration of strength this year. So thank you very much. I hope the presentation was informative.
I thank you for your support of the company, as shareholders. And I'm very pleased to deliver another set of strong numbers to you. So Chris, back to you.
Thank Thank you, Mark. Well, meeting is now open for questions general questions on the business. We did receive a number of questions in advance of the meeting from shareholders. Thank you, all those who did submit the questions. They can be grouped around some themes.
There's questions on remuneration, and we'll put those a bit later in the meeting when we look at the remuneration report and Bill Pulver will be speaking. There are a number of questions on the COVID-nineteen pandemic and the effect on the business. But I think Mark has sought to answer those within his presentation. There are a few specific questions which were received before the meeting, which we can handle now. And of course, there may be other questions coming in shortly.
So going through, but Mark and I can look at these questions now. Will Appen offer a dividend reinvestment plan to allow shareholders an opportunity to invest the dividends back into this great business? The answer is we don't have any plans at this stage for a DRP. When will you have a rights issue? So that's another thing.
We don't have any plans for a rights issue at this stage. Is there any incentive to provide a dividend increase in the near future? I think as Mark said in his one of his last slides there, we're looking at the overall capital management of the business. We're in uncertain times with the pandemic. We want to keep a strong balance sheet.
So the dividend policy will be caught within that general discussion, but it is a live issue for us to review it now. We have a number of questions from the Australian Shareholders Association, Mary Curran. Thank you, Mary, for those. Some are one question is REM related, which we'll leave for later on. There is a I guess it's a comment here from Mary, which I'll read out.
During the financial year, there's been a need for Appen to raise to capital raise and issue more shares. We understand this is part of the plan. However, I would ask in future for fair percentage for retail shareholders. Whereas I appreciate timing can be critical, the ratio offered to retail was not consistent with the register and the scale backs in the past to retail were 9070% approximately scale backs. Retail shareholders have grown from 8,000 to approximately 18,000 in this financial year.
Yes, look, we're sensitive to that, Mary. For the Figure eight acquisition, I'm sure, as most people would realize, it was based around a negotiated price with the vendor. We needed certainty. We needed to raise capital to deal with that. It was time critical.
So that drives us towards the dominant part of the capital raise being a placement, which was underwritten. That was successful. Because we wanted to reward loyal shareholders, retail shareholders, we did offer a share purchase plan, which was well subscribed and there was some scale back for that. But look, it's an ongoing issue, how to balance the company's capital needs, the critical timing of that if it's acquisition related and support for retail shareholders. We're sensitive to that.
Another question coming in. So I stable, when I said the update stable, our software tools are arguably world leading. Why use the word arguably? Is there something equal or better out there? What competitive business exists that is arguably also world leading?
Just wanting a sense of why you aren't more confident here. I guess I'm cautious. These things are absolute. Frankly, we do believe we are the largest provider and the most technologically sophisticated. But let me ask Mark if you'd like to expand on that.
Yes. Thanks, Chris. And I think it's a little safer to say arguably because there's some pretty smart people in our space. There's work going on in universities around the world that may result tools that are better than ours. But I think what we can quite confidently say is we have the broadest set of tools.
The competitors that we bump into tend to be specialized in one area or the other, most often image work, sometimes speech work. Very rarely do they cover the whole range of data modalities that we work in. So our strength and something we're confident in saying is that we can cover more data types, high quality data because of the crowd that we use. And I'm with Chris. I would say, when we get to a specific area, we think we're right up there, but we don't know everything that's out there.
So it's not a lack of confidence. It's just I think our style and just being rational and a little conservative about it.
Thanks, Mark. Is there another question? This question, as you get bigger, the target to maintain your current growth rate becomes more challenging each year. At what point in the future do you see constraints on your organic growth rate? Mark?
Yes. Thanks, Chris. And it is a common question because we've grown quickly. And I think those of us that know us well, I may be the wrong person to ask because I've got it wrong in the past because we've grown quicker than we expected. But we're confident that we'll see double digit growth for some years.
As the business gets bigger, of course, the law of large numbers kicks in and those growth rates at some point won't be as high as they have been historically, but we're still confident in seeing double digit growth for some time. It's a high growth market. AI is moving in ways that there will be AI applications next year that we won't have foretold this year. So we feel that our investments into technology, which give breadth and give us productivity and scalability and we feel that our investments in sales and marketing will drive high growth rates for some years to come.
Another question. Can you comment on how the need for more data more often has improved or not the visibility of your forward revenue? Can you characterize that as similar to recurring revenue? Mark, have a go at that.
Yes. Thanks, Chris. So as the business grows and as our customer base grows and as the work with those customer with those customers continues year on year. Yes, we do start to see repeat patterns in the data requirements, particularly when it comes to relevance data, which as I said, ages very quickly and so it needs to be refreshed constantly. Does this is this recurring revenue?
We hesitate to use that word because it has a connotation from an accounting and software company perspective. It's definitely repeat. And you can see in the cohort chart in the presentation that, almost all of our customer cohorts provide revenue, year after year after year. So the Figure eight platform gives us the opportunity to get customers to commit to annual revenue amounts. And of course, we're trying to grow that amount of committed revenue year on year.
But overall, there is an ongoing need for data, as the longer we work with our clients, the better we get at predicting that and the higher the quality of our revenue and earnings year on year. Thank you.
Okay. What impact has there been or is currently on Appen, if any, from the scare last year in which many of Appen's key largest customers had leaks from contractors of Siri, Hey Google, Alexa, of course, have private conversations.
Mark? Yes. Thanks, Chris, and thanks for the question. This is a really important issue. Data privacy around the world is something that we absolutely respect and we comply with all the legislation, things like GPR in Europe, for example.
Our clients, of course, have to comply with data privacy as well. And if we work with them, we have to comply with their standards. In the instances that there have been leaks, I'm confident in saying that the work that we're doing hasn't been behind any of those. I think that goes to the strength of our systems and also the care that we take around these areas. Perhaps the question is, well, would that require your customers to do something different and could impact your business?
I think to the first part of that, the answer is yes. They have to be more robust around the security of the data and the way they collect data and the way they use data. The second part of the question is it actually plays to our business, because we have a number of features and a number of capabilities that are highly secure compared to our competitors. So for example, while the majority of work goes to an at home crowd, which arguably is less secure, we do have facilities in The Philippines, in China and The UK, and we're implementing one in The U. S, where we get our workers to come into a facility.
Those facilities are accredited for what's called ISO 27,001, which is an information security standard. The workers are sort of scanned as they go in. They can't take phones. They can't take memory sticks, the computers they work on a lockdown, etcetera. And we work in those facilities to high levels of security.
So that's one capability we have, and we do a lot of work with our customers in that regard. Of course, in the current pandemic, that's been impacted, and customers are allowing us to do the work at home. But under normal circumstances, that's a real strength of our business. We also have, a secure browser based product that, improves the security of the at home worker. So it's a product that, locks down the user's computer.
They can't load data. They can't, unload delete data. If they were to take a photo of the screen, it would have their names or watermark over the top of it, etcetera. And so that gives us another level of security for at home work. So overall, data privacy and data security is really important and we have some highly competitive capabilities that mean we can participate in that work even at the highest levels of security.
Okay, another live question. News yesterday that Apple has just bought a small AI business. Is this a sign they'll take some of Appen's work in house? Do you see any other signs of this happening from any of your customers? It's a pretty lively M and A world out there, Mark.
Yes. And our customers are extremely voracious when it comes to picking up fresh and new technology. I think to the in this particular instance, and I think I'm aware of the company that was acquired, to the best of my knowledge, doesn't impact the work we do with our clients. In general, our clients do some data labeling in house regardless. It may be small pilot projects.
It may be, something that they don't think they want to commercialize. For example, they may have data that's been previously labeled that's useful for the particular use case. Typically though, when the customer gets to, a production decision, they want to take a product into production and they go from thousands of data points to hundreds of millions of data points, that's when they come to us. And that's what we're really good at, is that very large scale, very high quality, very quick provision of that data. So our customers will also do always do some work themselves, but when they need the data at scale, they come to us.
There's also there's a couple of other factors that go into why is there any reason our customer would not take such an operation in house. First of all, it's not their core business. They outsource a lot of stuff. They don't build their own computers, for example. They buy them they rent data center space, etcetera.
So we're a very outsourceable part of their business, albeit an important one. But second of all, because we're dealing with tens, hundreds of thousands of contractors, it's a really there's challenging elements to our business to running all those contractors and treating them fairly and legally. And typically, it's not something that our customers are built to do or want to do or have the expertise to do. So I think overall when our customers get to the scale of requirements to build production products, they need to come to us. And secondly, there are elements to our business that are just things they just don't want to deal with and don't have the skills to deal with.
So I hope that helps with that question. Thank you.
Next question. With the Figure eight acquisition, you're seeing much business from the big cloud platforms like Amazon SageMaker, etcetera?
Yes. So we do have partnerships with the large tech platforms. And the partnerships are built on their machine learning practices and their customers needing data labeled. And we've got, in some cases, the Figure eight platform is what I call white labeled. So if a customer goes to an online platform for machine learning infrastructure and they need data, they end up through on the Figure eight platform.
It is, however, early days for those platforms. I think it's early days not for the work just for the work we're doing with them, but for those platforms themselves. But nonetheless, we have existing partnerships with those businesses. And in time, we believe they'll be an important part of the business.
Okay. We've got a few questions to go, so let's try and keep the answers succinct if we can. Can you comment on your pricing models in terms of what and how much priced based on usage or data size versus price based on hours worked to annotate the data?
Yes. Thanks, Chris. So we do price in two ways. We price on the amount of data that the customer buys and sometimes we price on the hours of work that the customer does for us. Roughly speaking, the big relevance programs can be based on hourly pricing and most of the speech and image and some of the other relevance work is on data provision.
So it varies. We are trying to provide extra value on a data point or volume based pricing mechanism, because as we accelerate the provision of data using machine learning, we will be able to provide the customer with more data. So that pricing model is advantageous to us and the customer. But the short story is, yes, we use both pricing models depending upon the customer requirements.
Given the current pandemic environment and the wide range of possible outcomes, have you had to be more conservative in terms of giving guidance? Well, that's a good question.
It certainly adds complexity, but we are providing the same methodologies. We are using the same methodologies for our forecasting and guidance as we've used in previous years. But we're not blind to the fact that it does add to the complexity of our decisions. And hence, we are doing things in the business such as holding on to cash, given the uncertainty around the complexity. But the forecast is using the same methodology.
Regarding your marketing initiatives, what type and size of companies do you plan to target?
So the majority of our revenue many of you know comes from the large global tech companies. So we're looking at companies in a number of segments, Fortune 500 or enterprise companies, the largest banks and retailers, etcetera. Companies that fall outside of the 400, but have a need for high level training data, companies that are AI specific, for example. We have sales teams in Europe, Asia Pacific, of course China and the government. So we are targeting many different segments, to add to our customer base and diversify our revenue.
With the growing presence of TikTok in the social landscape, is there a move to engage or provide effort services to that customer?
So our team is actively engaged with all of the big Chinese tech companies, and we're pretty pleased with the progress they're making there. Many of you that have followed us will know that the work we do for our clients is sometimes on forward leading product development. So our clients ask that we don't mention them by name. But I can say that we are doing a lot of work with the big tech companies in China.
Okay. We'll take two more questions. Do you see other areas of business within the AI space that you could naturally migrate into the future? So we do look
at other areas such as, for example, a broader set of tools to contribute to the development of AI products. But we're really good at collecting data and really good at labeling that data and the demand for high quality training data just continues to escalate. So whilst we look for other opportunities, we've got to make sure that we don't miss out on the opportunity that we're banging the center of.
Okay, so could you please provide some more update in addition to ahead of plan to profitability on Figure eight. Is the figure eight customer base more in the SME, hence more likely to be negatively impacted by COVID? Any update on the ARR and customer churn in the last couple of months would be appreciated. So Mark, I'm not sure what we can say about that, but I'll leave it to you.
Yes, thank you. So the person that asked the question is right. Some of our smaller customers are more at risk during the pandemic than our largest customers, and we're watching that very closely. We're still confident that the path to profitability ahead of schedule. Of course, that comes from, not just managing revenue, but also managing cost.
Going forward, we don't intend to report on figure eight as a separate entity. It's fully integrated. It's very hard for us now to unpick the revenue and the cost of that entity. But I think the person that asked the question is correct. We do have to keep our eye on those smaller customers.
They are more at risk during the pandemic.
Okay. So I'll make this the last question. And any others, I guess, we can take on notice and get back to the questioner. The question, the crowd, where is the crowd? How much are the crowd people paid?
And on what basis?
Yes, thanks for the question. So the crowd is scattered across 130 countries around the world. The largest cohort is in The U. S. That's where we do
So that's where we have the largest cohort of crowd workers, but they are spread across 130 countries. They work from home. They probably work in coffee shops. They probably work on regardless of whether they're providing data that we charge the customer per volume or per hour. We pay the crowd worker an equivalent hourly rate that is commensurate with the work they do for us and in all cases above the minimum wage.
So as part of our crowd code of ethics and our fair work our fair pay policy, we've set pay scales around the world that are above minimum wage, in some case, well above minimum wage to get the skills we need. And we make sure that even if they're working for five minutes, they're getting an hourly equivalent that is consistent with that fair pay regime. So the crowd is scattered around the world and they're all paid fairly. Thanks for the question. Chris, back to
you. Thank you, Mark. We'll give you a short break now and move on to the formal part of the meeting. The notice of meeting was sent to all registered shareholders within the notice period required, and I take the notice convening this meeting as read. Voting on resolutions.
Before moving to the various resolutions to be considered today, I'll briefly outline procedures to today's meeting. In accordance with the company's constitution and as set out in the notice of meeting, voting on all resolutions will be decided on a poll, which I now declare open. Also in accordance with the company's constitution, the directors have determined that each shareholder who is entitled to vote at this meeting is entitled to a direct vote on a resolution. We've adopted these procedures to ensure that the views of as many shareholders as possible are represented at the meeting. Shareholders will be able to cast their vote using the electronic voting card received when online registration was validated.
If you have any questions about casting your vote online, please refer to the virtual annual meeting online portal guide or use the helpline specified on the screen in front of you. So as normal, the results of the poll will be declared and released to the ASX shortly after the close of the meeting. I've been advised that all proxies received for the meeting have been checked, and I declare them valid for voting. I'll disclose the proxy votes on your screen prior to the vote being taken for each item. These figures are as of the closing time for receipt of proxies, which was ten a.
M. On Wednesday of this week Wednesday, the twenty seventh. There are a number of voting exclusions that apply to the resolutions being put to today's meeting. These were outlined in the notice of meeting. So the resolution in Item two, remuneration report, is a nonbinding resolution.
All other resolutions are ordinary resolutions, meaning that to pass, they require more than 50% of votes cast by shareholders to be in favor of the resolution. If sufficient votes in favor of the resolutions are received, they will come into effect. As the Chair of the meeting and as detailed in the notice of meeting, I will vote where authorized all undirected proxies in favor of each resolution. So Item one of the meeting is the financial statements and reports. The first item of notified business is to receive and consider the financial report, the directors' report and the auditor's report for the year ending thirty one December twenty nineteen.
There's no formal resolution required for this item, but I invite questions and comments. So are there any questions or comments on the financial report or the reports of directors and auditors? Questions for the auditor in relation to the conduct of audit, the preparation and content of the auditor's report, the accounting policies adopted by the company in relation to the preparation of the financial statements or the independence of the auditor in relation to the conduct of the audit. No questions received on this. We do have Tony Nymak from KPMG available for questions.
We've got Kevin Levine, the CFO also available if there are any questions. But I see none, so we'll take that item as complete. Thank you. Item two is the remuneration report, the next item of business. So there are some questions on that which have come in and group those together and try to answer them through Bill's report.
So before the vote on it, the issue, I will ask Bill Pulver, who's the Chair of the Nomination and Remuneration Committee to report. Bill?
Thank you, Chris, and good morning, everybody. It is my pleasure to introduce the remuneration report, and I thank all of those investors and proxy advisers who engaged with me over recent to discuss their issues and questions. At Appen, we seek to ensure that our remuneration plans are closely aligned with the best interests of shareholders, but also fair and appropriate for attracting and retaining high caliber senior executives. Our process involves the use of external industry research on Australian and U. S.
Compensation trends. In Australia, our industry cohort is based on ASX listed companies with market capitalization between 50200% of Appen's market capitalization. We target our total remuneration levels between the fiftieth and seventy fifth percentile. Our senior executive compensation is heavily weighted towards performance based pay to align with shareholder interests. In 2020, our CEO's short term incentive plan will represent 21% of his total remuneration and his long term incentive plan represents 57%.
The combination of STI and LTI results in 78% of his total remuneration being at risk. The STI is measured against simple financial metrics of revenue and underlying EBITDA, the latter as our best proxy for cash flow. The LTI is measured against underlying net profit per share with a hurdle rate of 20% growth each year. The difference between underlying and statutory being the impact of the acquisition of Figure eight. For Australian based executives, the LTI has dual vesting requirements.
If the 20% hurdle rate is achieved, it vests in that year. However, the executive is also required to be employed at the end of the three year period for any payout. This way our plan delivers the retention impact we're seeking with our senior executives. For The U. S.
Executives, the LTI also includes a 20% hurdle rate. However, it vests annually, which is typical of U. S. Remuneration practices. The most common adverse feedback we receive on remuneration is an objection to the retesting we allow on the LTI.
If our CEO misses the 20% growth hurdle in year one, we allow him to recover it in year two if he delivers 20% times 20%, meaning 44% growth. The retesting is only achieved if the cumulative targets are met, which means over our three year plan, the CEO must deliver 72.8% growth in underlying net profit per share to achieve the full payout. In the event that the CEO misses his target in year one, we do not want him to stop at 20% growth in year two, which is behavior that a plan without retesting would encourage. Ladies and gentlemen, we are delighted with the performance of our senior executives who have delivered investors strong total shareholder returns over the last five years. Therefore, I hope we can look forward to your support with the appropriate resolutions.
I will now deal with the questions we've received online in advance of the meeting, but please note some of these have already been covered in my previous remarks. In relation to our resolution to increase the pool of funds available to pay director fees, please note that this increase is to accommodate the appointment of an additional director, taking our Board from six to seven directors. The fees payable to those directors in 2020 are unchanged from those paid in 2019. Please also note that directors do not receive bonus payments. There were several questions asking whether it is appropriate to freeze or cut compensation given challenges associated with COVID-nineteen.
Please note that management has reconfirmed our guidance for the fiscal year guidance, which was provided prior to the outbreak of COVID-nineteen. However, please also note that management is continuously reviewing cost management opportunities for improving productivity. Finally, to the Australian Shareholder Association's question about how we report senior executive compensation in the annual report. And I note that they are requesting an additional summary table on actual compensation, which we will consider for next year's report. I also note, thank you, the positive feedback from the ASA on the quality of this year's report.
Thank you, ladies and gentlemen. So Chris, now back to you.
Thanks, Bill. We'll move to questions that have come in this real time in a second. But the resolution before the meeting is to adopt the remuneration report for the year ended December 3139. So I think there is a question there on the screen. Bill, if you could handle that.
Yes, question one. Can you please confirm the 2020 target for LTI is 20% underlying earnings per share? Can you please disclose the 2020 revenue and EBITDA targets for STI? In relate yes, I can confirm that it's a 20% growth hurdle again for 2020 and the following two years. In relation to disclosing targets, Mark has given guidance in relation to EBITDA.
I don't believe he's given full guidance on revenue. So I'll leave those I'll leave any further discussion on the financial targets to Mark. But certainly, the EBITDA target has been disclosed. Thank you.
I think
that's the only additional question we've got. So now can we move to the votes received, the proxy votes and the votes received to date on the screen? Thank you very much. So now shareholders, please select either for, against or abstain for Resolution one on your electronic voting card in relation to this item. I'll pause for a minute or so to allow voting to occur.
Okay, so we'll move to the next item, which is item three on the notice paper, election of Director, Ms. Vanessa Liu. This resolution is in relation to the election of the newly appointed non executive director, Vanessa Liu. As I mentioned earlier, we invited Vanessa to join the Board March of this year and she's based in New York and brings several valuable dimensions in skills and experience to our Board, in particular technology, U. S.
Markets and global perspectives. So the resolution to the meeting is that Ms. Vanessa Liu being a Director who was appointed by the Board on twenty seven March twenty twenty and whose appointment as a Director expires at the conclusion of the Annual General Meeting of the company in accordance with Clause 60 of the company's constitution and being eligible, office herself for election elected reelected as a director of the company. So before opening this item for discussion, I'm delighted to ask to invite Vanessa to say a few words about herself. Vanessa, we'll switch to New York.
Thank you so much, Chris. Hello, everyone. Greetings from New York, and I'm very excited to join the Board of Appen after having spent many years in technology. I just wanted to introduce myself. I first started my career at McKinsey, where I was one of the leaders of the media practice, primarily in Europe, in Amsterdam, and London, and then ended up in New York.
And then after that, I became an entrepreneur. And based in New York, I started a venture startup studio where I cofounded two businesses in the media and commerce space. And since 2018, I've joined SAP where I lead SAP's early stage venture arm accelerator activities in North America, where we look at early stage technology companies in artificial intelligence, in Industry four point zero, basically wherever there's digital transformation. And I'm very excited to bring those skills to bear here at Athens. So thank you so much for your support, and I look very much to working together.
Thanks, Vanessa. So we're now open for any questions that might come in. I see none. On the screen, you can see the proxy results that have come in so far. And now the item is open for shareholders at this meeting to vote.
Please select either for, against or abstain for Resolution two on your electronic voting card. And we'll pause for a few seconds to allow that voting to occur. Thank you very much. Now for the next item, it relates to my own reelection, and I'll pass the chairmanship of
the meeting over to Bill. Bill Pulver. Thank you, Chris. I put the resolution to the meeting that Mr. Christopher Von Willer, being a director who is retiring in accordance with Clause 68 of the company's constitution and ASX Listing Rule 14.4 and being eligible, offers himself for reelection be reelected as a director of the company.
Before opening this item for discussion, I would like to ask Chris to say a few words about himself.
Thanks, Bill. And look, I'm grateful for the support which I've had to date. And I'm very pleased since our IPO to have shared the success of the company with you as my fellow shareholders. I'm a cofounder of Appen. I've been associated with the company since the very early days and of course, pre the IPO, post the IPO.
And I've been very excited about the growth and the history that's occurred. My main contribution I think I can make is relation to both the technology area. I'm an engineer by background. I've been in this advanced technology area for a number of years. I had a number of years' senior executive positions in Telstra before Appen.
I was Chairman of the Warren Centre for Advanced Engineering for several years at Sydney University. So I've been in high-tech and related areas for a while. I'm currently a member of the Australian Government's Digital Experts Advisory Committee, which is an interesting area to contribute to. And I'm an early stage tech investor in some emerging companies in satellites, in robotics and in agricultural technology. So this area is certainly a passion for me and I'm keen to contribute.
I've only taken one other board role, which is in Pacific Broadband Satellites Limited. The dominant part of my professional activities is related to Appen, and I feel I can contribute further. The Appen Board has a good spectrum of skills. And my contribution, I guess, is in the tech area plus the corporate history of the company. I work very well with the CEO, Mark Brayan.
We've got a very productive working relationship as I do with my fellow directors. And I'm pleased to offer myself for reelection. Thank you.
You, Chris. I now open this item for discussion. And I can see on the screen at this point, there are no questions. Details of the votes received for this item are on the screen. Please now select either for, against or abstain for Resolution three on your electronic voting card.
Thank you, and I'll now hand it back to Chris.
Thanks, Bill. And please stay in the chair there in case of any questions on the next item. The next item, Item five, the grant of performance rights to Mr. Mark Brayan, Managing Director and Chief Executive Officer. The resolution in front of the meeting is that for the purposes of ASX Listing Rule ten-fourteen and for all other purposes, shareholders approved the grant and issue of 78,125 performance rights to Mr.
Mark Brayan, the Chief Executive Officer and Managing Director of the company, and the subsequent issue of shares on the vesting of such performance rights on the basis set out in the explanatory notes to this notice of meeting. So I now open this item for discussion. I note that Bill in the REM report that he issued earlier touched on the basis of the LTI and performance rights, how they're structured. So are there any other questions? I see from the screen that no questions have been submitted online.
So we can now move, I think, to the proxy results there on the screen. Come in. And shareholders in this meeting can now vote. Please select either for, against or abstain for Resolution four on your electronic voting card. Thanks very much.
The last item, Item six, is the nonexecutive directors' remuneration in relation to pool of remuneration available for nonexecutive directors. And again, the background, I think, to why we're doing this was laid out by Bill earlier in the meeting. So the resolution is that for the purposes of Listing Rule ten-seventeen and for all other purposes, effective from the close of the meeting, the total amount that may be paid in aggregate and in any one year by the company to the nonexecutive directors as remuneration for service be increased by $100,000 from $800,000 to $900,000 Any questions? There's no questions coming in, so we can now proceed to the proxy votes received. We put those on the screen.
There we are. And this meeting can now vote. Please select either for, against or abstain for Resolution five on your electronic voting card. Thank you very much. Well, ladies and gentlemen, this concludes the formalities of the meeting.
I hope that the technology around the various shareholder groups has been successful. It certainly seems to be working at this end. Any shareholders that haven't submitted their votes should do so now, please. The poll will remain open for a further few minutes to allow you to complete the voting on your electronic voting card. We'll just give that a minute or two.
I now declare the poll closed. As I mentioned earlier, the results of this meeting will be announced to the ASX as soon as they've been counted and verified. I now declare the meeting closed. I'd like to take this chance to thank my fellow directors and Mark for and his management team for their diligence and commitment to this business. I'd also like to thank you, the shareholders, for your support and for your participation today.
I look forward to meeting with you again at next year's AGM. Thank you.