Tuya Inc.

Good morning and good evening, ladies and gentlemen. Thank you for standing by and welcome to Tuya Inc's fourth quarter 2024 earnings conference call. At this time, all participants are in the listen-only mode. After the speaker's presentation, there will be a question and answer session. To ask a question during the session, you will need to press star 11 on your telephone. You will then hear an automated message advising your hand is raised. To withdraw your question, please press star 11 again. Please be advised that today's conference is being recorded. I will now turn the call over to your first speaker today, Mr. Raj Chai, Investor Relations Director of Tuya. Please go ahead, Raj.

Thank you. Hello, everyone. Welcome to our fourth quarter 2024 earnings call. Joining us today are founder and CEO of Tuya, Mr. Jerry Wang, and our co-founder and CFO, Ms. Alex Yang. The fourth quarter 2024 financial results and webcast of this conference call are available at ir.tuya.com. A replay of this call will also be available on our IR website in a few hours. Before we continue, I refer you to our safe harbor statement in our earnings announcement, which applies to this call, as we will make forward-looking statements. With that, I will now turn the call to our founder and CEO, Mr. Jerry Wang. Jerry will deliver his remarks in Chinese, which will be followed by corresponding English translation. So, Jerry, please.

Hello, everyone. Thank you for attending the Q4 and 2024 business development conference. In 2024, we achieved nearly 30% of revenue growth and two major financial milestones. One is the first-ever non-GAAP business quarter and annual profit. The other is the first-ever non-GAAP net profit. These two indicators proved the unique business model of Tuya, Hello, everyone.

Thank you for joining TUYA's first quarter 2024 and full year 2024 earnings score. In 2024, we achieved nearly 30% year-over-year revenue growth and reached two significant financial milestones First, our inaugural quarterly and annual non-GAAP operating profitability. And second, our first-ever annual GAAP net profit. The former validates Tuya's unique business model, while the latter provides us with greater flexibility at the statutory shareholder's equity level, enabling us to take further strategic actions. These achievements strengthen our confidence in the company's future growth.

2024年是 Tuya 堅定地投入 AI 研發的一年。 In Q2 of 2024, we released our own AI model, the space model. By continuing to operate on-device AI, we plan to greatly enhance AI smart product experience to bring better product competitiveness to customers, and to bring more valuable smart product experience and functions to users. We believe that AI, IoT, cloud, and other technologies will drive new user experience and truly enhance the integration rate of device smartness.

2024 was the year of unwavering commitment to AI research and development. In the second quarter of 2024, we launched Tuya's proprietary AI large model, spatial large language model, while continuing to advance on-device AI. Our goal is to leverage GenAI to significantly enhance smart product experience, providing our customers with superior product competitiveness and end users with more valuable smart features and functionalities. We firmly believe that AI, IoT, and cloud technologies will drive a new era of user experience and substantially increase the penetration of smart devices.

Over the past year, we have firmly continued to upgrade the unique translation and design mode of the graphics platform, extending from PaaS to the solution mode of hardware products. Whether it is in the smart home or hotel, real estate, residential, travel, and new employees, we will continue to provide competitive hardware products for our developing customers, so that customers can participate in the competition from the software competitiveness of Powered by Shoei to the competitiveness of the supply chain and hardware products. As a result, we have gained more customers' recognition, and our Smart Solution revenue has achieved about 58% of the total growth in the year.

Over the past year, we have almost made significant strides in upgrading Tuya's unique software and hardware integrated business model, extending beyond past into a hardware solution model. Whether in our core streams, smart home, or in commercial verticals such as hospitality, real estate, mobility, and renewable energy, we're dedicated to providing developers with increasingly competitive hardware solutions. This allows our customers to expand their competitiveness beyond powered by 2R software capabilities into supply chain and hardware advantages. As a result, we have gained further customer recognition, driving an approximately 58% ELB growth in our smart solution revenue stream.

At the end of 2025, we will do our best to build a global AIoT development ecosystem, including continuous relaying of our AI agents and 2R OS cloud products, Looking ahead to 2025, we will be fully committed to building the global AIoT developer ecosystem

This includes ongoing iterations of our AI agents and the Tuya OS developer products, as well as AIoT open source software solutions, such as Tuya Open, to build long-term competitive advantages. Meanwhile, we remain committed to capitalizing on growth opportunities in international markets, accelerating the development of AIoT and AI applications worldwide, and delivering long-term value to our shareholders.

To allow more time for discussion and Q&A, our co-founder and CFO Alex will now provide more details on our business and financial performance.

Hello, everyone. I'm Alex. Today, I'd like to share an overview about our 2024 performance and our thoughts for the future. Before that, let me summarize the recent financial results. We wrapped up Q4 2024 on a strong note, delivering about 82 million in revenue, representing a solid 27.4% year-over-year growth. Given a higher base for the prior year, this growth demonstrates strong momentum. Our Q4 past revenue reached about 59.3 million US dollars, up 25.7% year-over-year, driven by our strong foundational market position and value proposition, enabling rapid transmission of customer expansion and end market demand into revenue. SAS and others' revenue was about $11.5 million, up 21.1% year-over-year, benefiting from the stable growth of SAS value-added services across our extensive devices space. Smart solution revenue reached about $11.3 million, growth by 45.5 year-over-year, supported by robust data demand across categories such as gateways, central controls, and energy efficiency solutions. So for the entire year, our total revenue reached about $298.6 million, making nearly 30% year-over-year growth. our overall gross margin remains stable at around 47%, while annual operation expenses declined by approximately 10% year-over-year due to the continually cost discipline of the company. As a result, revenue and gross profit growth efficiently translated into operation profit, leading to a 7.4% non-gap operating margin and a 25.2% non-GAAP net profit margin, with GAAP profitability achieved for the first time as well. On the cash flow side, we generated around $80.4 million in positive operating cash flow for the 2024. Following the dividend declared in August and pay in October, we ended Q4 with a net cash balance of over $1 billion. maintaining a strong liquidity position. Throughout 2024, we observed sustained growth in end-market demand, stable customer relationships and their increased investment in a smart product line, and the rapid adaptation and adoption of GNI and LLMs. So this drives a positive response from the consumer electronics industries towards intelligence trends. For Tuya, our unwavering commitment to the developer platform model combined with our unique software and hardware integrated approach and customer-centric product expansion strategy. So that has amplified those industry tailwinds, helping both of us and our customers achieve solid business results. In 2024, the number of IoT PaaS premium customers growth by 11% to 298. Our revenue dollar expansion rate, so DBNER, exceeded 122% at end of the Q4, marking five consecutive quarters above 100% and three consecutive quarters above 120%. Our products and services offering provide customers with all necessary technology and software to efficiently grade competitive smart devices and bring them to market. So the DBNER matrix reflects our continued ability to enhance our customers' competitiveness through 2S technology leadership. As a B2B company, our top 10% revenue contributing customers maintained retention rate as about 97%. underscoring a strong thickness of our neutral, open, and scalable technology platform. So our strategy, strategically focused on key accounts, has driven efficiency improvements too. Evidently by about 37% and 40% year-over-year increase in average revenue per customers and average gross profit per customers. 47% increase in company-wide revenue per employee. So at the same time, the scalability of Tuya ecosystem, our strong market positions, has freed us from rigid marketing and sales experience, allowing us to dynamically allocate resources, strategically promote new products, capabilities, feature set, and precisely target key markets. and expand our influence as well. Consequently, our sales and marketing efficiency improves by about 40% year over year. In the past business, 2024 witnessed a more diverse and dynamic developer base. So contributing to solid every year growth across all categories. Our product categories structures has become increasingly balanced and diversified. align well with the trend of customer expansion their product lines. In a highly fragmented and dispersed global consumer electronic market, those widespread adoption of technology continues to drive innovations across diverse product categories. This increasing variety of smart devices is fostering a more comprehensive approach to spatial intelligence. enhancing user convenience and comfort, ultimately progressing towards the vision of an interconnected and interoperable intelligence ecosystem. We're also pleased to see that our smart solution business model efficiently aligns with the need of top-tier customers, achieving about 58% year-over-year growth in 2024. For instance, in response to the French government's Energy subtype policy was seized and helped our French customers in becoming among the first to meet the country's energy efficiency subsidy standard. So this incentive program is set to helping millions of French households to achieve green energy savings. At this stage of smart technology development, Whether in consumer electronics or in industry-specific applications, a rich ecosystem of smart devices is essential for top-tier customers. In this regard, Tuya possessed a significant advantage over other players throughout its expansive developer ecosystem. As smart solutions further enhanced our ability to support customers, strengthen engagement, and boost the market competitiveness. And in the same time, we remain committed to building a robust developer ecosystem. At the end of 2024, the number of registered developers on our platform reached around 1.32 million, with over 1.3 million 1.07 million SQUs of smart devices developed onto our platform, spanning more than 3,000 product categories. We continue to foster an extensive ecosystem by, for example, integrating with Google Home APIs to create seamless smart home experience and collaborating with Cherry to establish a new vertical home interoperability ecosystem. At the same time, we are dedicated to expanding to a global influence, positioning ourselves as a reliable partner for the customers and developers worldwide. Our HEMS, the Home Energy Management System Solutions, has recognized in the United Nations Global Compact Report 20 best corporate sustainability practices in 20 years, highlighting Tuya's commitment to sustainability. Additionally, we achieved MSCI ESG rating of A and WINDS ESG rating of A2, and we were included in the S&P Sustainability Yearbook China Edition, and so on. So next, I'd like to discuss Tuya's opportunity in device and edge AI. As a technology-driven company, Tuya fully embraced the GMAI and the LAMs in early 2023 In Q4 of 2024, we launched the Tuya AI agent development platform, integrating all major large language models, including ChatGPT, QWin, DeepSeq, Dobao, LateChat of Mistrial, Gemini, Amazon Nova, and Cloud, et cetera. So Tuya's LMM agnostic approach eliminates the complexity of developing smart devices and applications from the ground up. providing developers with a crucial middleware layer that bridges LLAM's capabilities with real-world applications. Those developers, they can flexibly choose the most suitable AI models based on their business and marketing needs, while leveraging Tuya's templates, library, or customized solutions to develop AI devices and applications. This philosophy, aligned with Tuya's cloud agnostic strategy, ensuring customers do not have to worry about compatibility and sustainability of cloud services, IaaS, device types, or hardware architectures. Those agnostic compatibility provide 2D developers with unparalleled flexibility and adaptability, distinguishing us from single category solutions. Moving forward, We'll continue to prioritize AI devices and spatial intelligence applications, focusing on areas such as audio-video interactions, efficiency optimization, and decision-making automations. Audio-video AI enhanced user devices interaction by enabling advanced content input and output mechanisms, while efficiency and decision-making AI helping end users to optimize their smart devices and usage strategy to meet personalized and differentiated needs. We also continue to explore the application of Tuya's spatial LLAMs in energy management and other spatial intelligence scenarios as well. 2025 will mark the breakthrough year of device and edge AI. We plan to integrate AI capability across all categories within the Tuya developer platform, ensuring that every powered by Tuya devices is AI enabled by default in the future. together with a global developer partner, will explore various innovations and scalable applications for AI device, continuously shaping a variable global AI IoT developer ecosystem. Looking ahead, intelligence will evolve through the integration of software and hardware. Construction differentiated scenarios throughout interoperable smart devices and meets the customized needs of individuals, households, and spaces. So industry remain in penetration-driven phase with a vast and promising total addressable market. To achieve sustained growth, we'll focus on the following key areas. The first, we'll continue to expand the global market penetration. We'll leverage a combination of past SaaS and smart solutions to deepen our growth reach. particularly in Europe, Latin America, and the Asia-Pacific region, while increasing customer use cases. For the large enterprise customers, we offer Kube. Our smart provides private cloud solutions to help them to build secure and scalable enterprise-level smart platforms. Second, advancing the AI device and applications, we're committed to leverage GMAI and LAMs to significantly enhance smart product experience, driven product competitiveness, and delivering great values to the end users. We'll continue to innovate across high potential markets such as smart companionship, smart outdoors, smart energy, and smart spaces, complementing and accelerating smart devices penetration by AI applications. Third, throughout our integrated hardware software smart solutions, we are helping top-tier customers to speed up their product launches and establish differentiated competitiveness advantages across different categories and regions, directly increase end-market penetration, and delivering a greater and more sustained value. Fourth, maintaining a customer-centric approach. We aim to serve high-quality core customers efficiently, supporting their business growth, increasing customer stickiness, and repeated purchases, and ultimately enhancing our operation average and efficiency. Last, to continue to build our global developer ecosystem, we continue to iterate and refine 2D's AI agent platform, empowering developers worldwide to create customized AI devices and scenario-based applications easily. Additionally, we'll leverage GMAI tools to improve the development phases efficiency as well. So finally, let's address some frequently asked topics from the capital market regarding to the internal operation efficiency, the share-based composition expenses, and dividends. Throughout the 2024, our total headcount has remained at about 1,450 employees, reflecting on about 12% reduction from 2023. And by Q4 of 2024, our average revenue per employee has exceeded early 2022 levels by more than three times. Over 70% of our team consistently consists of R&D, technology, development, and a product person, so who drive rapidly the product integration across 2ES business line and prepare for the next generation of opportunities. Notably, 2ES revenue growth is not dependent on heavy investment in sales and marketing, which has been key factors in our stable and continuous improving profitability. We're committed to further improving the operation average while our net profit structures and quantity will become even more sustainable and organic. Regarding share-based composition expenses, the current quarterly accounting expenses primarily stem from the legacy granted and issued at our ESOC in a higher evaluation years ago. which has become amortized in quarterly, according to the vesting schedules. So those expenses are unrelated to any recent equity grants. As those legacy awards are gradually fully vested, SBC expenses will see a substantial decline starting in 2025. leading to a visible reduction in the accounting impact in our income statement, an improvement that's already evident in Q4 of 2024. So earlier this day, ahead of this earnings call, our board of directors approved the second dividend for 2024, totaling about $37 million, given our robust non-GAAP margin. Normally, non-GAAP matrix reflects the direct results of operational decision making, excluding external facts unrelated to our business model. So those non-GAAP margin and strong financial positions is the base we offer for the dividend, with over $1 billion in net cash and seven consecutive quarters of positive operating cash flow from Q2 of 2023 till now. We believe that TIA is well positioned in sustain a long-term growth throughout its competitiveness mode. And while also rewarding our shareholders who have demonstrated steadfast support, we remain committed to driven forces in both our global business and capital markets. Overall, 2024 has been a profitable year for Tuya, marking our first year of operational profitability, the execution of our AI strategy, and a breakthrough in our shareholder structure. During this strategic upgrade, we have been fortunate to experience multiple dimensions of progress and transformations, whether in achieving operational profitability, expanding our global footprint, or optimizing our international shareholder base. Notably, our partner with 65 VP of subsidiary of Tematic, the strategic investors, has positioned them as Tuya's largest institutional shareholder. Furthermore, our achievement has been strongly recognized by the market in early 2025. I believe this success is rooted in Tuya's global presence its unique hardware and software integrated business model, and its strategic focus on the global developer platform. Thank you, everyone. Operator, we can now begin with the Q&A.

Thank you. We will now begin the question and answer session. As a reminder, to ask a question, please press star 11 on your telephone and wait for your name to be announced. To withdraw your question, please press star 11 again. Please stand by while we compile the Q&A roster. We will now take our first question from the line of Timothy Chow from Goldman Sachs. Please go ahead, Timothy.

Thank you. Thank you, Megumin, for taking my question, and congrats on the very strong to end the year of 2020 school, and I think your presentation is very helpful. I have two questions here. One is that I think AI impact in your smart solution business could measurement further elaborate on what kind of usage scenario that you are seeing.

Yeah, thank you for the question. So for the AI, what we see here is that this will be a really good year for the AI. And we'll consider it as the first year that we can really turn the AI devices into reality. And what we see the priority of the AI opportunities on the device side. So we'll start with all the audio and video interaction devices, which means that to bring a new way to have the people interact through the audio and video to the device. So that would be one thing. So no matter is some use cases where they show that in the last quarters, like the pet feeders, like the bird feeders, and the PET related appliances and some control panels on the wall with a screen as well. So that will be one thing. And the second thing is that all those types of analytic type of devices that need a complex decision. So one significant use case is energy solution. Because for comprehensive energy solutions, you need to be able to read through dozens of variables from different types of devices and from the outside factors, like the weather, like the utility data, et cetera. And then make some adjustment automatically or very dramatically, making new decisions to operate the device in different ways. So the AI definitely will help in that part. So that'll be something. And also what we see is a very interesting new type of stuff will be all type of toys. So we found that the kids will be the perfect adapter for AI because they don't have any stereotype about any existing technology, so they grab anything as a fashion new, so they just easily tap in. And so since then, January, after the CES, we really got plenty of customers reach out to us, bringing different ideas that how they can integrate AI into toys. So that would be some of the, what I see, incubating categories we started catching up. But I believe there will be more booming stars throughout the entire year, or even in the next couple of years. While the AI, starting from the very fundamental stuff, like the LAM, someone can bring an easier approach to the market to activate all the new ideas. And so while the barrier become low and you have thousands of thousands of company and the developer join in, so that will become a robusted market of what we're looking forward to do. So there might be thousands of different type of things coming after.

That's it. Well, Timothy, do you have any follow-up questions? Sure. Can you hear me well? Yes, we can. Please proceed.

Great, great. Sorry about that. Yeah, my second question is regarding your margins. Just wondering, I think, with the AI enhancing the demand for IoT, what is the unit price trend over time and how that will impact your margin? I think it's more on the gross margin level. And can you also talk about your OPEX, given I think you have been investing on AI and operating efficiency on the OPEX level has been quite high. Just wondering how do you think about your operating leverage into 2025? Thank you.

Yeah, so the first one is that the AI will be still in a very early stage. So we're not making a final decision or we didn't make up the final call about how we're going to commercialize the AI. So we tried different type of commercial use cases. And in one word is that we're starting to deploy the AI. into all three business models we have, including the PaaS, the SaaS, and the solutions. So we offer the AI in different ways based on what type of needs it is. So that would be one thing. And so sometimes it will be a new offering that we provide to the market. So we will repricing that as a new product. So for that part, The margin will depend on not only the cost, but also depends on the demand. This depends on what type of reasonable pricing we like to put into the market to scale it. And another part is that some of the AI, if we think that will be very, very, we want to put that as the default features that by enhance all customers to improve their competitiveness. So we'll integrate that into our existing path solution too. So the margin impact, I think that in the short term, I don't have visibility for that because the scale might not be growing that significantly to influence our overall margin. So that will be smaller pie at this moment. And in the long run, I believe that AI will bring more values. But we will not be committed to when. Because for sure that the focus for us, I mean the priority for us is always scalability. Yeah, that would be one thing. And the second thing about OPEX, so like I mentioned earlier, we're starting to invest in the AI once the GMAI occurs back to early of 2023. So it's not a new investment for us throughout the past two years. We're ready to relocate our resources internally and improve our talent structures in some roles to be capable of building AI already. So right now, there are a lot of departments in my company that are all AI ready now. So which means that I don't have to recruit a team to rebuild things all over again. It's already been there. And so for this year, we will slightly increase some of the specific application and investment by the demand of the market slightly. And also in the same time, we're looking forward to see what will be the right opportunity or right type of way that we can market it in a better efficiency, since it comes with a new concept. So you better make some noise. But that type of impact, we'll see that it's very controllable. So I'll say that we're not improve the OPEX significantly. It will be in a very manageable level.

Thank you. That's very helpful. Thank you.

We will now take our next question from the line of John Roy from WTR. Please ask your question, John.

Can you hear me okay? Yeah, I can hear you, John. Great. Excellent year last year. Very well done. So I was wondering about SAS and its growth prospects. Do you see that as something that could really change the landscape, or is that just more of an add-on?

Yeah, that's a good question. Thank you for that. So the first one is that SAS is based on the base of the total deployment of our devices. later business models versus the past, which means that we need to enlarge and scale the past, and on top of that we can do more SAS. That would be one thing. And second thing is that we're facing structural improvement in the SAS model, which means that we try to increase the portion of the recurring models in the SaaS type, not only the software base, but also the recurring software base. So that would be the second thing. And so for that part, I'm not worried about the growth. I'll pay more attention on the thickness of the customers, whether they can use that type of thing. I don't know if they can use that. So the payment, the continually, annually payment will be a very strong base for us. That would be the second thing. And the third thing is that Within the past, so there will be some, one portion of that is for the commercial software. So it's not the consumer-facing one. So for the commercial one, in the past three years, the major market we're trying to do this business now is in Asia and China. So because that's within COVID period, so which means that it brings us difficulty to be able to incorporate new business across the global basis. And starting from late last year, we're starting to duplicate data use cases in different countries. We try to bring that into global market. We really have some really positive progress there, no matter in Latin America, Southeast Asia, and in Europe too. So for that part, right now we are approaching to a bigger total addressable market on the regional basis. And we've already found the duplicatable applications or products that are proving to work out. And we already have a customer base that's applying with recurring models. So for the SAS, I believe that will be our long-term growing business line. We're not looking forward to any short-term surprising or short-term significant improvement, but for the non-real, maybe after five years, that will be a very, very joyous thing for us.

Great. Thank you for that. So, obviously, the dividend is very nice, and you have a pretty significant cash position. What is your view on the acquisition front? Is that something you're actively watching? How are the prices in that market?

I think that... I'm thinking, you mentioned about MMA, right? MMA is always an open option for us and so we keep screaming and what type of extension, what type of companies or partners can be our extension, either to extend our scope of coverage of different types of scenarios or to be a more vertical solution provider for specific industries. So we keep screaming on that. Yeah, so I think that's an option for us. I think the key part is that the first one is that no matter with or without acquisition tool, those guys should be our developers. So the first priority is to have a very strong global developer ecosystem. I think that will be our major focus. In the same time, anyone maybe within the ecosystem that can show a better potential, no matter, like I mentioned, it's that horizontally extend our scope or vertically that improve our vertical industry insights. So for that part, we can easily identify that. Yeah. We'll keep that open. And we'll keep for anyone to help us to find some interesting target too. Do you have anyone introduced to me?

Will do. Thank you so much. And again, congratulations. Thanks.

Thank you. We will now take questions from the line of Kai Hsiao from CICC. Please ask your question, Kai.

Okay. Then the second question is about our big customers. I saw that the high-quality PaaS customers under the big customer strategy of the company have a very strong growth. I would like to ask how the company will further expand the resume of high-quality PaaS customers in the future. And if we master some of the layout in GNI, is it helpful to help the company to expand more high-end customers? There are two questions in total. Then I will translate it here. So this is Ben from CIS Research and congratulations. for the strong quarter. So my first question is regarding the IoT path. So what's the current downstream demand for IoT path and which countries and categories you think have more potential and how can we outlook the growth margin? And the sixth question is regarding the high quality customers. So we see the number of customers have very strong growth. So what do you think the further potential for the company to further expand the high-quality PaaS customers in the future? And will GenAI help in the acquisition of more high-quality customers? Thank you.

Thank you for the question. So the first one for the PaaS, as we can find in our business, it's very strong showing that we have a very balanced structure on the business. So no matter is on the regional side or on the category side. So right now the business comes from over 3,000 categories on the 2F platform. So we didn't rely on any single category. So which means that we're quite diversified in the category side. So that's what I mean that whenever or whatever type of new booming stars in the smart devices market occurs, for sure that there's someone making that into your platform, into your ecosystem, not from somewhere else. So for sure that we will cover any new opportunities. And on the regional side, right now, we're very balanced too. So the largest regional market in Europe only covers slightly over one third of 35 of our business on the end user demand. And for all other countries, North America, Latin America, Asia, and Australia, including China domestic, so they're quite balanced. So we didn't rely on any single region market, too. And which bring us a very good barrier or very good protections for any type of change happening on the right now, the geographic scenarios. So that would be one thing for the past. And for the second question, for the second question, I think the major customer here is that right now, for the first 10 years of the establishment of this company, and the larger portion of the ecosystem come from the consumer field. So what we found here is that two extensions on the customer side. The first one is that for the past three years, we have more and more customers and use cases come from non-consumer fields, no matter it's commercial one, industrial one, and that type of things. And also, in the company, we have new business lines just focused on that part. So that'll be one thing. That will be total extension and some new type of customers from that part, including the telecom carriers. So we got, we continue to have new telecom carrier partners coming in. So that'll be one thing. And the second thing is that exceeding from the device-based stuff, in the near future we believe there'll be more and more spatial solutions being needed And the special solution providers usually be a different type of company too. And for that, no matter is for the property manager, warehouse manager, logistic. So those can manage some type of space. And the house, the building, the warehousing, or some asset too. So those type of solutions, before AI technology happens, they don't have a comprehensive write-up solution to carry the pain point. And while the spatial models and spatial solutions we design can meet that, so that can help us to open another door, too. So that will be two, what we call the new for us.

That's it. Thank you. Kai, do you have any follow-up questions? That's all. Thank you.

Thank you. Our next question comes from the line of Yang Liu from Morgan Stanley. Please ask your question, Yang.

Thank you for the opportunity to ask questions, two questions here. The first one is on the outlook of the revenue structure because we observed in the past few quarters that IoT SaaS and other technologies growth recovered meaningfully and while the smart device distribution growth has been relatively volatile. So this kind of exchange is very critical to the overall growth margin. So I would like to hear your view in terms of the three baseline relative growth rate going forward and whether it will have meaningful impact on the blended cross-module. That is my first question. And my second question is regarding what's your view on the computation in the AI for the IoT solution provider. Do you see more hyperscalers who would like to develop IoT hardware themselves, or you think it is still or is it a booming market with a lot of smaller product development companies in the market that can continue to add a lot of value to those customers? Thank you.

So for the question, I think that our three business models, because they come in with a totally different approach to the customers. So I mean, overall, the margin level of each of the business models will be kind of stable. So for the PaaS, we believe that the overall margin for PaaS will remain at around 45%. We think that will be reasonable, or slightly over 45%. And for the SaaS, because it's software-based, I think that over 70% will be remained. And on the solution side, because we'll follow the hardware game, So while you have, you scale the volumes, while the customer scales the volume too, and sometimes you have to subsidize some of the profit to the customer. So we believe that on the solution side, maybe by we scale it into, you know, we double that, or we scale that in a bigger base, that maybe the margin of solutions will slightly decline a little bit, but not that much. We think that we'll continue to maintain the solutions margins over 2021. So that'll be overall the margin, or the nature of the margins, underneath each of the business models. And put that together, and what we found is that the paths will continue to provide a stable growing base for the entire two-year business, because that will be essentially everything that the customer needs to purchase. And SAS, like I mentioned earlier to John's question, so we think that the SAS will be more long-term growing curve for that. So we're not in a hurry. So it continues to enlarge the device base and the customer base, and we deploy those services on top of that and convert more of them to pay the recurring revenue. That will be a long-term journey, but remaining similar growth with the past, maybe that'll be our expected to. And solution, because solution come from my past customers, they calling out for solution rather than the development toolkit from the past time. So solution is that we convert to more and more customers' demands directly from my past base. So for that part, the solution might go faster. And while solution become a larger pie of my total revenue, And so my over general growth margin rate, growth margin, my GPM, we believe that in the long term will slightly decrease a little bit. But I think that would be reasonable. I mean the margin would, because I have a bigger pie of the solution side, would come with a lower margin compared with the other two business models. So for that part, We think that would be good. But that's not the problem for us. I believe that the key to evaluate the gross margin is that we need to break down each of the business model and tell the value of that. Whether that values reflect to the right margins as that business model, and I need to compare that business model to the industry level. So the SAS is compared to other SAS company or other SAS and the path is farther past it. So that's the point I have to try to make. So it's a very stable business model. And the structure will start to change slightly.

Thank you.

Thank you. How about the second question? What is the outlook for the competitive landscape in AIU for the IoT?

Yeah I think that, I think I really covered part of the question. I think that right now the good part for AI is that right now it's in a very positive and active momentum for the market, not for capital market, but I mean for the hardware industry. And so we have a lot of movers as either designers or manufacturers or solution providers and so Right now, we prioritize three. Like I mentioned, it's either the video-audio interactive, or the decision-making analytic one, and the toy. But we believe that there will be thousands of, at least thousands of more vertical use cases we're coming from. So it's very early stage, and this year is that we got so many players on the field to try different type of ideas. So what we need to do is that we offer them a toolkit to easily turn their ideas into two. I think that will be the unique value that Tuya provide. And also that's why we have such a successful developer ecosystem. So to let them get away from do everything in-house, we offer them a shortcut to testify their ideas. I believe that this year will be kind of very starting line for the AI device that you can find millions of new devices out there. A lot of people try in and some will prove working out and starting to scale that. Maybe some not. They redesign that. So, yeah, I think that will be the momentum we're looking forward to see. And so what we do is that to just, like I mentioned, continue to lower the barrier for anyone trying to get in.

Thank you.

Thank you. As a reminder, to ask a question, please press star 1 1. There are no additional questions at this time. I'll hand back to the management team for any closing remarks.

OK, so thank you again for joining our call tonight and this morning. If you have any further questions, please feel free to contact us or request through our website. We look forward to speaking with everyone in our next earnings call, so have a good day. Thank you.

Thank you for your participation in today's conference. This does conclude the program. You may now disconnect.