Okay, let's begin. Good afternoon, everyone. Welcome to Plover Bay 2021 annual results earnings call. Today's call will be presented in English, and all currency figures are in US dollars unless otherwise stated. For today's presentation, we have Alex, our company's founder, and myself, Chris, the company's CFO speaking. Before we begin, I just want to remind you that the results has been published on hkexnews.hk. We will also share with you the slides later on. Right. Let me jump in to introduce our financial performance in the past year, and then Alex will share with you our business growth and opportunities going forward. Next page. Next page. First, let me summarize the financial numbers for 2021.
In 2021, our revenue increased to about $74 million, an increase of 40%. Our net profits increased to $21.2 million, which is a jump of 49% compared to last year. During the year, our gross margin stayed relatively stable at 58.2%, while our strong operating leverage has helped expand our net profit margin to close to 29%. We recorded a diluted earnings per share of $0.0194, and that's an increase of 46% year-over-year. Finally, we declare a combined second interim dividend and a special dividend of HKD 0.0898. Next page. Diving deeper into different product segments, in 2021, we recorded quite a strong growth in all segments and geographical regions.
In terms of product segments, our wired SD-WAN segment increased 36% year-over-year, while our wireless SD-WAN segment revenue increased 50% year-over-year. The strong increase can be attributed to strong market demand for small branch, ad hoc networks and remote networks. These kind of networks serve homes, offices, shops, vehicles, ships, events, and many other applications. We see that the need for easy setup, reliable connectivity is growing everywhere. Besides market demand, we also did particularly well with our 5G products. We believe we currently have the most comprehensive 5G product portfolio in the market, and 5G products accounted for 10% of our revenue. Keep in mind, that does not mean our 4G products are slowing down. In fact, our 4G products are also benefiting from the overall strong market demand growth, as I mentioned just now.
Next, our warranty and support services revenue increased 28% year-over-year. That's driven by a growing user base, which will support our future growth of this segment. Then lastly, our software licenses segment grew 16% year-over-year, driven by quite a strong increase in the subscriptions for our InControl 2 cloud-based management platform. I also like to mention that the number of registered device in InControl 2 has grown 30% year-over-year to over 350,000 devices now. Next, in terms of geographical breakdown. Revenue from North America grew 46% year-over-year. EMEA grew 36% year-over-year. Asia Pacific grew 20% year-over-year. Other regions grew over 100%. The growth, again, is very well distributed across our key channel partners and reflects a strong underlying market demand for our products.
Of note is that in Others segment, we began working with a new distributor for Australia and New Zealand back in March last year. Last year, they helped us broaden the number of certified products that we can sell in the ANZ region. This market has been in an initial stage of ramp up last year and is expected to contribute a lot of growth in the coming years. Next is a discussion on recurring revenue. Our recurring revenue increased 29% year on year and accounts for 23% of our total sales. It may seem that the proportion of our recurring revenue has dipped slightly compared to around 25% in the last couple of years. I will explain that our recurring sales typically lags our router sales by around one year.
This is because when we sell a router, we have to defer a portion of sales over the coming, next 12 months to reflect the free one-year warranty and services. Then also, customers don't get to resubscribe after a year. Looking at the deferred revenue in the balance sheet, our total deferred revenue figure increased 33% during the year. Most of this balance will be booked in our revenue in 2022. Our newly launched paid services, such as SpeedFusion Cloud, SpeedFusion Connect, these will also grow from a low base to contribute to our growing recurring revenue stream. Next, on gross margin. Our overall gross margin is 58.2% this year, which remains largely unchanged from last year.
Overall, we see slight increases in most of the segments, as you can see here. This is offset by our mix, product mix continuing to shift towards wireless SD-WAN routers, which has lower gross margin compared to our overall company's gross margin. Therefore, our overall gross margin has stayed pretty much flat. We're also pleased to report that our net profit margin has continued to climb, closing into 29% in 2021. During the year, we continued to see strong operating leverage. Our operating expenses increased 19% during the year compared to our 40% increase in revenue. Next, our balance sheet and operating cash flows remain healthy. During the year, we invested quite a lot in inventories to keep our product availability and lead times at a very reasonable level.
Given the industry-wide shortage of semiconductor components, inventory increased to HKD 18.6 million. Inventory days increased to 183 days. 70% of the inventory is actually raw materials, such as IC chips, integrated circuit chips and wireless communication modules. We will continue to invest working capital to ensure we have sufficient supplies to meet customer demand during the ongoing shortage. In conclusion, Plover Bay has outperformed other peers in our industry. Whether it is our revenue growth, our margins, or expenses ratio, this reflects our R&D-focused nature instead of having large sales organizations. We hope to continue to be able to deliver this kind of results to our investors. Next, I will pass the floor to Alex to talk about what we will do in the coming years and our business outlook.
Thank you, Chris. First of all, our revenue growth is higher than our industry peers as our focus is on fast-growing wireless first connectivity. We have one of the most, if not the most comprehensive product offering of the LTE, 5G, SD-WAN product portfolio in the market. This product are purpose-designed for various vertical markets. Our proprietary SpeedFusion technology ensures reliable connectivity over a fixed network. LTE, 5G or even Low Earth Orbit satellite. Combined together, we have the ability to reach many market verticals where others cannot. We also have a couple of structural advantages over our industry peers. Plover Bay is an R&D-focused organization with an outsourced sales team. We rely on our channel partners to scale the sales organization in their home country, home market. Products are priced to deliver strong value propositions to customers.
Our small teams in various global locations are able to rapidly respond to customer feedback and market changes. You can see our gross margin level is very similar to the industry average, but with our unconventional R&D-focused and outsourced sales approach, we have lower OpEx and a much higher operating margin. This model has been proven since our reinception. This has proven to be scalable and sustainable too. It is not easy to replicate as many industry peers already have a deep-rooted culture and organizational structure. Next page, please. Now, let's talk about the key growth drivers in 2021. We have seen a very strong growth in small-sized ad hoc networks around the world. Businesses are connecting their branches utilizing LTE, 5G as the primary connection. This business want quick deployment, affordable pricing, and easy-to-use solutions.
Our full-featured small platform appliance is very competitive in the market. Last year, we mentioned the multi-year upgrade cycle to 5G for better speed and lower latency. In 2021, we launched eight new 5G models, and these products have contributed to around 10% of our total revenue. Again, Plover Bay has one of the most, if not the most comprehensive product offerings in the market. Together with our extensive 4G product portfolio, we can address the current and future needs for our customers around the world. SpeedFusion is our iconic technology. It can magically bond multiple network links together, no matter if it is DSL, cable broadband, fiber broadband, LTE, 5G, or even satellite. They don't need to be coming from the same network operator. SpeedFusion can bond them together and become one big pipe.
The end result is higher reliability, higher speed, and especially the upload speed needed for two-way video calls, video surveillance, and remote works. The reliability enables mass deployment of wireless first connectivity. In 2021, we launched the SpeedFusion Cloud service to further simplify the deployment. Customers can just subscribe to the service without the need to set up their own SpeedFusion infrastructure. Our enhanced marketing effort has been very encouraging. Our dynamic and creative team has put their passion into action and created some very successful marketing campaigns in an unconventionally cost-efficient way. This is helping us to attract more customers as well as channel partners.
We have also migrated the Peplink Community to in-house developed software for better system integration. This is an important step for us as we could offer better user experience for our customers and channel partners to access the resources for customer support and future service subscriptions. Next slide, please. What about the future? The need for rapid deployment, affordable and easy-to-use networks are growing very strongly. We work very closely with a number of managed service providers with an entrepreneurial mindset to address this fast-growing market. In the last 12 months, we've won a good number of very influential projects in various vertical markets with our partners. We believe these influential projects will form the blueprint for future customers in these vertical markets over the next few years. This will be one of the areas of sustainable growth for us.
The cost per GB for both 5G and LTE has reached a new low level during the year. As we mentioned last year, while the fixed broadband has an unmetered fixed price nature, wireless WAN being a metered network has been a psychological barrier, keeping some organizations from deploying mobile broadband as the primary WAN. Now, this psychological barrier appears to be melting away quickly. We have seen many customers embracing multiple 5G LTE links to form an elastic dynamic on-demand wireless network, which is even more reliable than fixed broadband in certain locations. We will continue to expand our industry-leading pro-product portfolio as well. We are expanding our software ecosystem to grow recurring revenue streams as well. Over the years, we have developed the software building blocks internally.
We realized that by working with other software partners, we'll be able to deliver a better digital experience to simplify the connectivity and make it easy and more reliable. More features are brought to the table to retain and attract more subscriptions. Another interesting area is the emerging markets. We'll leverage our software capabilities and work with different system-on-chip and extend supply chain partners to expand in emerging countries. Through these partners, we will combine competitive pricing with stable software code and advanced features to create a compelling new product family for these emerging markets. While these markets may require a lower price point, but they represent a significant growth opportunities for us. Next slide. These are some of the application examples. On the enterprise SD-WAN side, our products serve the work from anywhere. Enterprise failover, business branch and cloud access, IoT and kiosks, mobility connectivity.
Managed service providers are using our products to offer unbreakable connectivity solutions. On the 5G mobile SD-WAN side, maritime, autonomous systems and teleoperation, specialized vehicles, aircraft, rapid deployment sites, and public safety are key areas of deployments. In summary, we are well positioned for growth in the years to come. Our addressable market is expanding. In addition to the enterprise edge networks, there are many new markets emerging from 5G, autonomous systems and IoT. Continuous innovation is at the heart of our team. We have a proven track record of innovations. We listen to our partners and customers for continuous improvement. You will keep hearing from us on new products and services to new markets. Expanding the ecosystem. We have an ecosystem of hardware and software as well as data plans. They are all managed by a single cloud platform.
While the industry is moving towards a SaaS model for everything and customers are expecting an engaging digital experience instead of just a point product, having a good integration of various elements in the ecosystem would require a very different mindset. Overall, we firmly believe we are at an early stage of a multi-year growth cycle. We are a passion-driven company. Problem-solving is in our DNA, and we have a track record. We have teams of energetic, committed members around the globe. We share one common goal of building something extraordinary, and we always choose to be great instead of big.
Okay, that's the end of the prepared remarks. We're open to the Q&A session now. If you have any questions, please feel free to use the Zoom's raise hand function to...
I'll unmute you, and then we'll get into the Q&A.
Hello?
Hello.
Hi. This is Vincent Fong. First of all, congratulations on strong results. Can Alex or Chris elaborate more about the geographic performance, especially given that we are hearing U.S. and other developed markets preparing for more opening up? What's the implication from this opening up of the economy to your business development in the first half of this year? Is it going to be more constructive or are you seeing a kind of tapering off given very strong growth last year? Entering into 2022, maybe the base is a bit high from last year. Thanks.
Hi, Vincent. This is Alex. I think as we're talking about is the supply and the demand. Definitely, we are seeing a very strong demand everywhere. In USA, in Europe, or even in Asia. I would say that we have no concern about the demand. At the same time is whether we are already having a very high growth or very large space or not. I think this is just the beginning. I do not feel like as though we are at a very high level of anything. However, on the supply side, there are many challenges. Say for example, as we all know, on the supply side, the semiconductor shortage. This thing is real, and this thing is really bothering the whole industry.
This is so for certain. Say for example, if we are missing some really tiny little components, that could delay the whole shipment. At the same time, the lockdown of various cities, potentially maybe. Of course, our manufacturing base is mostly in Taiwan. I would say, we still don't know about how the Omicron is affecting this production capability and things like that. At the same time, that the logistics is always still having a challenge. Luckily, the thing is, Plover Bay is actually still. We still operate as a very small company. We have a very hands-on mindset. This hands-on mindset is on everybody's mind. Not only on the
I mean, this is a broad mindset across every team members. Every people at the company are willing to take extra steps to solve the problems. That's why you can see even back in the Q4 or back in November, December, we already see there are a lot of constraint on shipping the products to USA. We are still able to ship a good number of products over there. I would say the demand side is very strong. Then on the supply side, we just need to keep having a problem-solving mindset in order to get that resolved. I think we are optimistic that we are going to keep a good growth pattern in the future.
We're not talking about a single quarter, a single year or anything like that, because again, we see all this growth is actually a macro sustainable growth. Because we are not really focusing on the large enterprise networks or anything like that. There are a lot of smaller networks, edge networks. These smaller edge networks, they demand the same reliability as the enterprise network. But at the same time is that this market cannot be met by the consumer products. On this market segment, we see ourself are in a very interesting and defensible position in this market, especially with our connectivity technology like SpeedFusion. Yeah, Vincent, I can-
Yeah, yeah.
Answer the questions.
Thanks for telling us that underlying demand is still very strong. Maybe two follow-up questions, if I may. Firstly, do you guys provide guidance for 1Q for first half? Secondly, since you have established a base in Lithuania, if I'm not wrong.
Yes, that's right. Yeah, that's correct.
Do you know what's going on? Any insights? Now Russia is kind of invading Ukraine.
You're referring to the whole thing, right?
Yeah, yeah. Will that have any indication to your component supply and maybe the whole industry? What's your take?
Okay. First question is, we do not provide any guidance on, in terms of numbers, at our current scale. 'Cause, again, we are here looking for long-term sustainable growth. We are still having the mindset that we will continue to do the right thing for sustainable growth for the company. That's for question number one. For question number two, what we really do in Lithuania is we have an operation team there, and we have an R&D team there. Basically we are not sourcing any components in Lithuania. Instead, we have our European warehouse over there, and then we have an R&D team there.
Of course, we don't know about the impact of the situation happening there yet. In the worst-case scenario, we just have to ship the product from Taiwan, from Hong Kong or even from other countries to Europe, if there's something happening over there. At the same time, for the R&D team, again, this is a human capital. Our team. I don't think so the human capital thing will impact that a lot, unless something really serious happening there. We just don't know at this moment.
Thank you.
Feel free to, jump in and ask a question.
Okay. I think we have a question from Ming Ran. His question is, what factors were driving the rising demand for the small ad hoc networks? I think there are a couple of areas. For example, some people replacing the POTS telephone line for Voice over IP. You guys probably have seen we have a case study with a pretty big coffee chain in USA. They are deploying our product for all their branches. The application is they are connecting these branches over LTE. That's one example. At the same time we are also seeing supermarkets or even grocery chains in Asia Pacific.
They are also moving to use mobile, the convenience store. Or even, we were seeing that in the last year. There are quite a lot of like the large food restaurant chains, they're migrating to use LTE or 5G networks. I think it's the fundamental reason is people want to save cost. People want to have a lower network cost, more flexibility to deploy the network. In situations like this, wireless is a better option. We see all these applications are pretty important. At the same time, in certain markets, we are seeing people are staying in the recreation vehicles. There's some people working from home or people can now work from remote.
That is when they work from remote, they might work from their recreation vehicle, the RV, or even in a campsite, or even in their vacation home. There are many, many remote locations. Now people are looking for reliability, connectivity over these remote networks. We also see the maritime market is growing very well because maybe now is people. Okay, there are actually two factors in the maritime segment. The number one thing is again the recreation vessels. It's people stay in the boat, and that they just need to get connected because they need to do the work over there. That there are also digitalization in the maritime industry.
Like, is the large container boats or those work vessels, they need connectivity too. In the past, their only option is just satellite. Now that they have a lot more options with LTE or even 5G for these applications. These are the driving forces for the rising demand for small ad hoc networks. Okay, I see there is another question from Brian Denyeau. On the LEO side, apart from Kymeta, is there any other opportunity that might be working with, say, Starlink's or Amazon's Project Kuiper, since the LEO technology is best complement to cellular? Mushroom Networks and Teltonika developed aggregation technologies similar to SpeedFusion. What's your view on the competition? From my understanding, IPv6 is still not supported for all current products, and it perhaps will be an industry standard very soon.
When would IPv6 be rolled out on your products? These three questions, let me answer them one by one. The number one thing with the Starlink or the LEO technology, we already supporting that. The thing is, when you're installing a Starlink, you have a modem. Starlink gives you a set of equipment that includes the antenna and a modem. You can just connect that modem to our products, to the WAN port of our products. We already do some software optimization to make the user experience easy. If you spend some time to our forum, you can see many Starlink customers are using our product to complement with cellular.
You're correct that the LEO is very complementary to cellular. At the same time, is this a competing technology to cellular? We don't see that that way. Because it's that we are seeing that people are looking for more connectivity options. No matter these connectivity options, be it satellite, LEO or LTE or 5G or whatever. The more, the better. Except the thing is, people might not want to pay a fixed rate for that, or they do not want to pay a fixed monthly rate for that. The on-demand piece is actually more welcome. That is our view on the LEO technology. Then it's also your second question about a potential competitor, Mushroom Networks and Teltonika. These are not new stuff.
First of all, is Mushroom Networks or other players. Yes, they are similar technology, aggregation technology similar to SpeedFusion. One thing that is, we are really proud of our technology is the installed base. Our installed base already tells what's the difference, because our product and our technology, it simply works. People want to use the technology, they want to experience the technology, but they do not want to troubleshoot with the technology. I would say is a SpeedFusion is probably still the most easier technology to use. In fact, we also have some customers, and these are actually pretty big tech companies. They told us that they spend time in developing a similar technology and yes, they made it. It works.
Why they still want to use our technology instead of their in-house developed thing is, they do not want to spend the time, the ongoing effort in going in solving this problem. They were telling us, "Hey, you guys solve the problem so easy, and it makes our life so easy. You guys solved the problem and make our life easy." That's the catch. They'd rather they're willing to pay for using the technology instead of maintaining that by themselves. I think is, I'm pretty sure that is there will be other people developing similar technology. Again, one thing is, the beauty is everybody's technology is proprietary. When you are having many proprietary to proprietary technology in the market, then is the installation base is important.
At the same time is we have built ecosystem to make life easier. Only is just one of the key thing why people love Peplink, and that is the original reason why people love Peplink. However, after SpeedFusion, we also have other things that is convenient to the market. For example, we have different hardware platform designed for different applications. If you want to put that into your robotics, if you want to put that into your small robot delivering for food delivery. Yes, we have a small form factor product there. At the same time is if you have a bigger vessel you need to connect, or is you have other teleoperation requirements, you want a bigger box. You want a high-end equipment. Yes, we have that platform too.
At the same time is the whole user experience and they're compatible with each other because they're all running on SpeedFusion. That means you can use InControl one dashboard to manage everything. You can manage that with your small little robot delivering the food for food delivery. You can even using the same dashboard, the InControl, to manage your larger network. I think this is actually the hard part for the competitors. The third question about the IPv6. Yes, IPv6 is again, this thing has been talked about for years. One day or I think I would say that we understand IPv6, this thing is needed. We are not sitting here doing nothing.
In fact, we have been working on the IPv6 for some time, but then for now, I can say that we'll never lose an opportunity because we are lacking of the IPv6 support. I would say IPv6, yes. We need to have that feature, and we are working on that, and we will have products supporting that. Brian, I hope I answered that, your question.
Question from Steven Wong.
Right. Chris, yeah, this one is for you.
Okay. How many months can our inventories be able to support the demand? I would say around three to six months. Right now it's closer to three months, but then we will have new inventories coming in. Large orders we placed last year, they will arrive in the second quarter this year. We're good for a while. Next question comes from Calvin Sham. Apart from the inventories, is there any bottleneck or maximum capacity in terms of production? Yeah. Let me take this question as well. We do not have our own production facility. We outsource all our production to contract manufacturers in Taiwan.
In terms of their total size, we are not a major player there, so I think we can ramp up easily if we see the demand. Again, back to Steven, another question from him. What's the progress in launching SpeedFusion Connect? What are we doing to help encourage customers to renew their subscriptions after the first year?
Okay. SpeedFusion Connect. Yeah. Okay, this is an interesting thing. We actually launched the SpeedFusion Connect earlier this month, so this is still in early stage. Actually, why we delayed the launch of this SpeedFusion Connect is we actually are fine-tuning the business model. Because we learned one thing is the demand for on-demand connectivity is actually larger than expected. But at the same time, there is no such solution in the market. We felt that, you know, in order to do this with the SpeedFusion Connect, this is. Actually, I should say is that we were struggling between should we do this by ourself directly with our existing channel partners. Should we position this product as a managed service provider or carrier solution?
We were struggling in the approach to address the market. I think our latest answer is these two markets do not contradict. We will be working with MSPs and operators to launch SpeedFusion Connect. Of course, the approach will be different. At the same time, in order to have SpeedFusion Connect, in order to see the magical effect of SpeedFusion Connect, we kind of require some form of hardware upgrade. This cycle we are very close to finish the upgrade cycle in order to support SpeedFusion Connect. That's the current progress. Again, we are very excited about that.
It's just we were kind of like evaluating the situation about should we do that by ourself, with our channel partners or with MSP. That's the status. The other question about is how to encourage customers to renew the subscription after the first year. This is actually coming from our software ecosystem. Again, with the SaaS, with our approach on the SaaS model, convenience is a big piece. Now we are adding a lot more interesting features. For example, when people are buying our product, then is they will have a free access to SpeedFusion Cloud for certain amount of data. Then is we even offer some form of a SpeedFusion Connect data plan to customers.
At the same time as we'll be offering a couple of new interesting applications, software applications to make connectivity easier. We believe all these convenience will just make customers to continue to renew their subscriptions.
Another question from Steven is, how is our progress in developing the direct sales business on Amazon? Okay. Let me take this question. This year our business with Amazon scaled up quite well. Right now it's around 2.5% of our business. It really grew quite strongly this year. We'll continue to deploy our direct sales on Amazon into the European market. One good thing about Amazon is because it's anonymous, it enables a lot of potential customers that want to deploy Peplink into their enterprise.
That they can just order from Amazon and get a delivery next day, and then they can put it into their network to test it out. This is one good thing about Amazon. Next question comes from Ming Run. With more hardware models supporting SpeedFusion Connect, can it choose the carrier based on signal strength on the location?
Yes, absolutely. Customer can choose the carrier based on signal strength, and customer can also choose the carrier based on the upper layer protocol. For example, we will be able to do certain health check on the networks. It can be based on the actual throughput of the network rather than just the signal strength as well.
Okay. Another question from Brian Denyeau. There are big names on the autonomous and teleoperation end users like Waymo or Toyota or Starship. What do you see in the coming years on those areas of deployments? Is it going to play a significant role on the growth? What's the competition on that area like?
Okay. First of all, some of these names are no strangers to us. We just can't disclose who the people using our products for autonomous and teleoperation are. What I can say is, in the autonomous and teleoperation industry, many people know about us, and they are very familiar with us. The reason why is the good thing is so many engineers, they have been using our products before. During their career development, I mean, during their career, they might change to different companies or they might work with different teams, they might work with different companies. They want to bring their experience and leverage the experience to work on something.
I would say that is definitely a key market for us and we have a significant role in the growth. The question is, when this is going to be material. We just don't know yet. The reason why is all these deployments. These are advanced deployments. If you are looking from a revenue perspective, the smaller network deployment. For example, the nationwide deployment of a coffee chain or the nationwide deployment of a convenience store. The revenue size from those businesses might be way, way bigger than this. Again, these two markets are not contradictory. That means we can do both. If we are looking from a revenue perspective, we just don't know when this is going to be significant.
However, we still see this is a very important market for us because to a certain extent, this is an endorsement of our technology and the advancement and the reliability of the technology. We still put a lot of effort in those markets. Again, if you understand our approach, we work with certain customers directly, and the reason for working with them directly is just because we learn from each other. Then, as always, we also collect good feedback from these direct customers. With this direct feedback from these customers, we are able to build better product, and we are able to customize or to gear our product toward our industry-specific applications. This is definitely influential and important for us.
Okay. There's a question from Benny Tang. What is Plover Bay current market share? Are you expected to have a larger market share in the coming years? Okay. I think right now, our market share is around 5%-7%. The market is the wireless gateway market. It's not very reliable to gauge our market share because we are expanding in many ways. For example, we besides selling SD-WAN routers, we're also getting into tapping into the data plan market. I think gauging market share is not the best way to look at our company. Brian asked another question. There were many M&A within the industry in the past few years.
Has Plover Bay ever been approached? What is your view if there is an attractive offer?
I think this question is very difficult to answer because we just don't know whether we can talk about that or not. What I would say is, if there is an attractive offer, definitely, we are open to explore the options. We are open-minded.
Feel free to shoot us a question. You can type it, type your message or unmute, and then just ask over the line. Okay. From Steven, do we have any medium-term targets for the company?
In terms of, Steven, actually, I think we always look for one thing, sustainable growth. Our mindset is always sustainable growth. I would say, in the near term or medium term, that is, we love to hire more people who wants to do something, who wants to achieve something. We love to hire people who are hungry and willing, whether they are really experienced or not, it doesn't matter. We really care about the mindset. That is our near-term thing. At the same time, the product development is always something we have working on.
The products, the services, that's always something we are working on. Do we have a financial target internally? Not exactly. Yeah, we do have objectives. At the same time these objectives are kind of like we would like to achieve the goal. Again, are we going to just solely focus on reaching that goal by ignoring anything else? No, that's not the case. We look into the whole business from a more holistic approach. We will be looking at what is the best and what is the most sustainable growth for us. That's the thing.
There's another question from Steven. How do we ensure the current culture can be maintained in the company as we grow larger? Do we see our R&D staff joining other competitors? Okay. How do we ensure the current culture can be maintained? I think we have a very rigorous. Okay.
Sorry, Chris. Yeah, because I was first seeing another question. I was seeing. Is that Timothy's question about the dividend policy? Let's answer the culture question first. Okay. I think that in order to maintain the culture as we grow larger and larger, our approach is actually we break into many small teams. We have a lot of smaller teams working together. Again, when we keep hiring, we keep growing. We still operate with a lot of small teams. Our hiring policy, we always try to find out the passion of the team member.
Whether his passion or her passion is on content creation, or the passion is on photography or anything like that. We really drive from that. Then is also with this kind of culture, then is the good thing is people get lot of job satisfaction. People also take strong ownership because we encourage experiments. We also acknowledge the fact that is that the more you do, then is the more mistake you might create. Then is but is we are fine with that. This kind of culture is actually very well received, especially among younger people. We believe is that we can scale this. 'Cause when we first started the company, we were only like is 10 people, and then is we scaled up to 200 something.
We believe that this can keep growing, because, again, we are not creating many, many layers in between. We just have a lot of small teams working together.
Going back to Timothy Wong's question, the dividend policy. How we decide the dividend, is it in terms of EPS or free cash flow? How much CapEx is required each year? Let me answer this question. Our dividend policy is basically, we want to reserve enough amounts for our working capital needs. For example, salary, rent, and also inventory investment. We do not set a very strict payout ratio each year. We will look at our cash balance and make a decision on whether there is a special dividend or not.
It also depends on whether we foresee there is a need for large investments such as CapEx or acquisitions. In short, we look at various factors, not just whether a percentage of our EPS or free cash flow. Hope that answers your question. Next is a question from Jeff Pang. How would you balance growing the developing market, which may give you lower margin while fulfilling the demand in the developed market given the current supply chain constraints?
Okay. Currently we are using primarily one SoC system on chip from one supplier, from one semiconductor vendor. With that semiconductor vendor, there's a supply chain going along with that. We are very happy with that. At the same time we are expanding to use other SoC players. With that SoC players, there is a different supply chain. That means we are expanding our supply chain to different SoC players. Contract manufacturers and everything is different. In situations like that, we are able to develop a different hardware product, but running the same software. Okay, our market is actually very interesting.
Because let's say for example, if we are selling into the U.S. carriers, certification is a big entry barrier. When you make a product, you need to certify that with Verizon, T-Mobile, AT&T. It is also actually quite a process. It's quite a very serious process. If we are selling to ASEAN countries, we might not need to have that level of certification requirements. In situations like that, we can work with a different SoC semiconductor vendors to address those markets. The semiconductor supply chain constraint, yes, it happens across the board. We believe that our approach is, we will use a different supply chain for the emerging markets.
Next is from Ming Ran again. With 5G's higher throughput, do you see more larger business customers using 5G or LTE as their backup networks?
Yes. Yes, absolutely. We actually have some very large customers. They are using multiple 5G for failover their fiber network. For example, we have hospitals in U.S., and then this is one of our case studies. They have multi-gig fiber. There's a lot of bandwidth on the fiber connection. They are using our product to bond multiple 5G connections, such that is if there's anything wrong with the fiber network, they can failover to the 5G. Yes, this is actually a key market for that for us. We also see in this market is we are probably the best player. I see we missed one question from Steven Wong. Do we see our R&D staff joining other competitors?
First of all, many of our R&D staff, they have been with our company for long, for years. There's a very long relationship together. They also grow this company. I mean, we grow this company together from nothing to date. We also believe we have a brighter future in the future. I would say a lot of time is if we just hire somebody, and then is so the first six months or the first year might be difficult for these people because we have a unique culture. If...
If these team members they stay with us for more than a year, I'm pretty confident that they will stay with the company. Because it's at the same time is that we offer a lot of unique values that it is difficult to find in other competitors or in other players in the industry. The question is, first of all, we don't see. Yeah, I think it's over the years. We've probably seen one or two, but then a lot of time is just because they might want to relocate to a different location. For example, maybe we have some colleagues, they just want to move to another location and then so that's a different story.
Other than that, our R&D staff is actually pretty stable.
Next, also from Steven. Can you share how the economics will look like if we are selling data? Are we taking a cut based on the data used? What would be the margin?
Okay. When we are selling the data, we are not going to undercut anybody because we are going to position this as a premium data product. Why this is a premium data plan is because first of all is that this is on demand. At the same time is that this can be longer period. Because now is when you go to buy the data plan, a lot of time this is a monthly plan, or a lot of time is they will expire in six months or a couple or something like that. We are going to package that completely different. Say, for example, is that this is becoming an annual plan or even multi-year plan, and then is, but this is still on demand.
Say, for example, if people are going to failover from the fiber connection to the cellular connection. They can just buy this data and put that then put the data plan sitting there until when the fiber is broken or when it is a Unison failover, then you just kick off that data, those data plans. Yeah, the answer is yes, definitely, we get a cut based on the data used. Then, in terms of the margin, we don't have the answer yet. Because as I said, we are working with different ecosystem partners to launch our product. Then I would say actually, the interesting thing here is not really about the margin. It's about the ecosystem. It's about the convenience.
We believe, if we have an ecosystem and we deliver the convenience, this is actually the biggest entry barrier for other people to go into this market.
From Timothy Wong. Various sources show that the SD-WAN industry should be growing around 20% per year. How come most competitors underperform?
Okay, let me answer this question. Actually, no, maybe you have confused it with the. No, we haven't mentioned that. Okay, how come most competitors underperform? I think you're referring to the comparables that we put on the slides. I think the SD-WAN is a very broad term. Most SD-WAN players are focusing on the fixed network, fixed line. They are not really focusing on the mobile SD-WAN or the wireless SD-WAN. In fact, we don't know any SD-WAN industry report referring to the wireless piece. Most of the SD-WAN piece is just fixed network first, the wired first approach. Probably this is not the accurate thing. I mean, in comparing these competitors with the SD-WAN industry growth.
It seems that there are no more questions. Well, thanks for everyone today. We had a very great turnout today, and thanks for everyone's support in the past year.