Huawei is “all in” on building a third mobile operating system and ecosystem. This is the first major challenge to the Android-Apple iOS duopoly in almost a decade. And as iOS is a closed system, it is really about Huawei taking on globally dominant, open source Android.
It’s a big deal.
But first…consider joining Asia Tech Strategy, my podcast and subscription newsletter on the strategies of China / Asia tech companies.
I recently spoke with Huawei about the current state of their operating system HarmonyOS. And it is a fascinating subject. I think it raises three really cool “strategy meets tech” questions.
- Can Huawei build a new operating system and developer ecosystem for smart devices?
- Can Huawei become an advanced software company?
- Can Chinese software finally move from the application layer to the system level (i.e., compilers, operating systems, programming languages, databases)?
This has implications far beyond Huawei.
When President Trump cut Huawei off from Android / Google Mobile Services in 2019 and 2020, it began the political weaponization of the tech supply chain. Reliance on the Android / iOS duopoly has become a risk for almost every smartphone and smart device maker in China / Asia (which is almost all of them). Companies like Xiaomi, Oppo, and OnePlus are now very aware that the US government can cripple them at any moment. And most of them do not have the financial resources of Huawei. Note: Xiaomi was just added to a different US blacklist in January.
Building alternative supply chains for critical tech is now a priority for everyone. And it is has been foisted on Huawei as a requirement.
Huawei Is Openly Taking on Android in Operating Systems. Finally.
For most of 2019-2020, Huawei was pretty dodgy about Android. Everyone knew high-end semiconductors and the Android operating system were the major problems in their tech supply chain. But their standard statement was that they wanted to continue working with Android. They were pretty quiet about their own operating system.
I figured they hoping for a reprieve from the entity list ban. And they were buying themselves time. Plus they likely had smartphones and other projects in process that still required Android. But it was obvious they were working on their own operating system as fast as possible behind the scenes. Even if they got a reprieve from the entity list ban, the risk would still remain. Developing their own operating system became a requirement regardless.
Well, in the last couple of months, Huawei has dropped the dodgy answers and is now openly pushing the roll-out of HarmonyOS. They are building an alternative to Android.
I routinely ask if I can talk with people at HarmonyOS and HiSilicon. Those are the two areas of Huawei I am most interested in. And, to my surprise, I recently got a yes on talking to HarmonyOS. So I spoke with Clement Wong, who is VP of Consumer Product Marketing. Note: there have also been some interesting comments recently from Wang Chenglu, the President of their Software Engineering Department of the Consumer Business Group. He appears to be leading R&D on the operating system project.
Based on all this, I have five take-aways on the current status of HarmonyOS:
- Take-Away 1: Huawei is Building an Operating System and Ecosystem From a Position of Advantage
- Take-Away 2: Huawei is Benefiting From Good Timing and Past Preparation
- Take-Away 3: HarmonyOS Is An Attempt to Leapfrog to a New Type of Operating System for 1+8+N Devices
- Take-Away 4: China Is Increasingly Going After System-Level Software
- Take-Away 5: It’s Ultimately About Winning Partner Trust and Participation
Take-Away 1: Huawei is Building an Operating System and Ecosystem From a Position of Advantage
Huawei is clearly copying the original Android approach in many ways. Google launched Android in response to the arrival of the iPhone and its new tech paradigm. Microsoft fumbled around. Blackberry / RIM ignored it. But Google nailed it. They:
- Launched the Open Handset Alliance, which brought together manufacturers, semiconductors and mobile carriers into an ecosystem.
- They made Android open source.
- They made Android free. However, they required a license for certain functions.
- They actively supported developers.
I discussed this in my podcast on ecosystem management.
- Alibaba, Android and The Emerging Art of Ecosystem Management. (Jeff’s Asia Tech Class – Podcast 57)
Huawei has done a lot of these same things.
- They have made HarmonyOS open source and free.
- They are encouraging adoption and usage by other handset and smart device makers. They have +20 hardware partners.
- They are supporting developers. They have 2.3M registered developers today. This is a mix of global and local developers. However, that is still small compared to the 20M and 24M developers for Android and iOS.
That’s all good. But it still leaves them trying to break into a mature market with entrenched competitors. That difficult situation is compounded by the chicken-and-egg problem. To get the developers, they need the consumers. But to get consumers, they need developers. Plus they also need hardware partners and other groups. There’s a reason why these ecosystems are so entrenched. And why the above strategy hasn’t worked for any other company.
But this is where Huawei is different. They have some advantages that other would-be operating system / ecosystem builders do not have.
First, they have hundreds of millions of consumers using their phones. They already have one of the user groups for their platform. Huawei has recently said that HarmonyOS will be installed on 200 million Huawei phones, and 30 million Huawei tablets, watches, TVs, and speakers this year.
If that big group has a good experience with HarmonyOS AND keeps buying their new devices, they will have secured one major user group. In theory, the developers should then come. That would be a big and proven market for their products. Investing the time and energy will make sense for them. If they get the developers, the other hardware manufacturers should start to adopt the system over time as well.
That’s the path forward. And, according to Wang Chenglu, they have studied the market and the key milestone is to get to 16% of the operating system market. That appears to be the magic number to get the operating system running. Pushing HarmonyOS out to their 200M users is the key first step in that.
Take-Away 2: Huawei is Benefiting From Good Timing and Past Preparation
The entity list was really bad luck. Nobody thought that was a possibility in 2016.
But Huawei has had some good luck as well.
- First, they have a massive consumer base, as mentioned.
- Second, there are also some important technological shifts happening in smart devices, which makes entering with a differentiated operating system easier. Timing is important.
- And third, they just happen to have an operating system launched a few years ago.
Keep in mind, Huawei’s smartphone and consumer business is a pretty new part of their business. It really only started growing fast around 2015. Yet in 2017, it had reached 20% of the China market. And in 2018, it was on the verge of becoming the number one smartphone globally (counting Oppo, Vivo and OnePlus separately).
It was during the 2016-2017 period that Wang Chenglu proposed to Ren Zhengfei an escalation of the operating system initiative that had been percolating within R&D. According to LatePost, he argued it was too much of a risk to be a leading smartphone maker and not control the core underlying technology (i.e., Android). Ren approved and the development of an operating system became a company-wide priority. Although it wasn’t exactly an urgent one.
Here is a summary of the history from Wang Chenglu in LatePost:
“Wang Chenglu: The project was officially initiated in May 2016 within the Software Engineering Department. This was possible thanks to Huawei’s R&D mechanism. Huawei R&D can be divided into three segments in a timeline. 2012 Labs focuses on the technical direction and trends we expect to see in the next five to ten years. The product research departments of each product line work based on a shorter time frame. They focus on the technical trends and specific technologies that will arrive in one to three years from now. The product development departments of each product line work on R&D to ensure delivery for the following year. In 2016, there were about 140 employees in the research department under the Software Engineering Department, which made decisions for the research team. We initiated a project to develop an OS prototype. By May 2017, Version 1.0 was released.”
“In May 2018, it received funding from the Huawei CBG Investment Review Board and became an official CBG project, taking it one step closer to commercial use.”
That was really good timing and preparation. In 2019, Huawei was cut off in high-end semiconductors and Google Mobile Services (GMS).
A Brief Aside on Innovation vs. Audience-Builder Platforms
I have been using the terms ecosystem, platform business model and operating system a lot in this article. I do define these things specifically.
I define ecosystems as endeavors that require the coordinated activities of many companies. These are projects that are beyond the ability of any one company. Ecosystems are pretty common in technology because software and hardware all have to work together.
Ecosystems are better at complex endeavors. But they are also less efficient than standalone companies. Traditional vertically-integrated companies (i.e., pipelines) are good at optimizing activities in a tightly controlled sequence (i.e., the value chain). Factories and their supply chains can be ridiculously efficient.
Ecosystems are also able to tackle more complicated and uncertain situations. They are more adaptable and innovative in the face of uncertainty and technological change. Their strength is that they can draw on the creativity, R&D and resources of many participants. You see ecosystems emerge at moments of technological disruption, when big challenges are presented and the path forward is not clear.
I define digital platforms as business models that are simple types of ecosystems. These are single companies that use business models that connect specific users groups. And enable them to do certain specific types of interactions. So marketplace platforms enable monetary transactions between merchants and consumers. Audience-builders enable the viewing of content creator videos.
I have outlined 5 types of platform business models (digital, not physical platforms). I find these to be the most common types. But others are emerging.
- Marketplace Platforms. Such as Alibaba, Meituan and Uber.
- Payment Platforms. Such as American Express and Alipay.
- Innovation and Audience-Builder Platforms. Such as YouTube and Microsoft Windows. More on this below.
- Collaboration, Coordination and Standardization Platforms. Such as Slack and Zoom. Lots of B2B and enterprise-stuff here.
- Learning Platforms. A complicated category.
Operating systems are both ecosystems and platform business models. They bring together lots of companies (hardware, software, semiconductors, etc.) but they also have one core interaction that is an Innovation Platform.
Innovation and Audience-Builder platforms are a complicated category. I basically consider audience-builders to be a sub-type of innovation platforms.
At the highest level, an Innovation Platform enables the creation and distribution of a digital good or service. It is something you can innovate and build something upon, usually intellectual property. This can be media, software programs, mobile apps, content, console games, etc. Microsoft Windows is an innovation platform that enables software developers to build programs that run on the platform. And that are used by personal computer users. Operating systems are innovation platform that have two main user groups: software developers and computer users.
I define Audience-Builders as a big sub-type of Innovation Platforms. This is about media, entertainment and generally building an audience around digital content or experiences. People making videos on YouTube and TikTok are innovating, creating IP and trying to build audiences. Spotify is connecting listeners with people making songs. Gaming is similar. Yes, it’s an innovation platform. But it’s mostly about building a consumer audience. And the whole idea of media and audience building is merging into social media, utilities and other activities.
Overall, I consider HarmonyOS an attempt to build an ecosystem AND a classic innovation platform.
Ok. Back to the topic.
Take-Away 3: HarmonyOS Is An Attempt to Leapfrog to a New Type of Operating System for 1+8+N Devices
So let’s say a mobile operating system is mostly an innovation platform, where the developers create an App Store full of applications. But it is also a broader ecosystem because it has to connect to the handset, the headphones and other peripherals. And it has to integrate with semiconductors and computer architecture. And, now that Moore’s Law is over, performance advances are coming from increasing specialization and software advances across the ecosystem. The days of easy gains based on advancing general CPU performance are gone.
So how do you break into a mature operating system and ecosystem with increasing specialization?
Ideally, you wait for a change in technology. You wait for a period of digital transformation or upgrading. That gives you the opportunity to differentiate. From Wang Chenglu:
“An OS’s success depends on whether or not it offers tremendous value to various players in the industry. If our products just keep doing the same thing, consumer experience stays the same, and the consumer may not buy them. In this case, developers do not have an incentive to develop applications, and the ecosystem will not experience healthy development.”
Fortunately, Huawei is trying to break into operating systems at a time when technology is evolving significantly. There is 5G, AI, cloud computing and edge computing. And Huawei has long been focused on the end-to-end process for communication. They do everything from smartphones and other smart devices to connectivity (5G) to cloud and AI. As mentioned, they were lucky with the timing.
For smart devices, they have long summarized their approach as “1+8+N”, which means:
- 1 smartphone as the primary device.
- 8 ancillary smart consumer devices that connect to the smartphone. Such as smart tvs, AI speakers, smart cars, etc.
- N (i.e., infinite) connected devices that connect everything to everything. Such as IoT sensors, video cameras, etc.
Their vision for smart devices is 1+8+N devices connecting to each other and working together. But each of these devices has very different performance and energy requirements. And very different cost structures. And right now, these devices effectively speak different languages which makes interactions difficult. Especially at the application level.
HarmonyOS’s approach is to create one integrated operating system for 1+8+N. For developers, it means an app written for HarmonyOS will work seamlessly on all types of devices. That’s the technological change and point of differentiation they appear to be targeting. It’s an operating system designed for 1+8+N. It’s both integrated and distributed.
That’s an interesting idea. It’s integrated like iOS. But it’s an open source system like Android. And it’s also distributed across device types, which is unique.
Keep in mind, HarmonyOS began as an operating system for IoT. For sensors, cameras and devices that would be everywhere and in everything. It began as software for tiny sensors with very narrow bandwidth, minimal computing capabilities and limited power. In 1+8+N, they started at N.
I asked Clement about how this plays out for consumers. Why is this better? He pointed to several scenarios that can be done across multiple devices.
- Smart homes.
- Productivity and smart offices.
- Health and exercise.
Ok. Two last short points.
Take-Away 4: China Is Increasingly Going After System-Level Software
Wang Chenglu spoke about this, which is bigger than just Huawei.
Basically, China is awesome at mobile apps. And Huawei and other companies are awesome at hardware and manufacturing. But China has never been a major player in foundational and system-level software. No operating systems. No database software. No programming languages. So what Huawei is doing is a significant move in this regard.
Clement spoke about the challenge of doing foundational software. That this is a long haul. You can win fast with an app. But to convince developers and hardware partners to build on your foundation takes time. Building an operating system that is reliable and trusted is a long haul.
Huawei is pretty well-positioned in this regard. They are already a major player in hardware and software. They are already working on everything from semiconductors to devices to connectivity to cloud. And most of what they have been selling for the past 30 years is infrastructure that others build upon.
Final Take-Away: It’s Ultimately About Winning Partner Trust and Participation
Everyone knows Huawei can build software and hardware. And nobody really doubted that their +194,000 employees couldn’t code a new operating system if they focused on it.
But to go from coding to an ecosystem, they need partners. They will need participation, Trust and participation are really the key metric for HarmonyOS. And I think 2021 is going to give us a solid answer on that.
Ok. That is for today. I will be getting back to public company strategy stuff. I just thought this was fascinating.
I write and speak about digital China and Asia’s latest tech trends.
I also run Asia Tech Strategy, a podcast and subscription newsletter on the strategies of China / Asia tech companies.
My subscription newsletter offers:
Deeper insights into the strategies of the tech giants of China / Asia. I help investors see around the corner – both with tech giants and rising companies.
See the big picture. Get a better understanding of Asia’s digital ecosystem. What are the important tech themes? What will the future look like? Where to hunt for opportunities?
A unique view from on the ground of digital China / Asia.
Note: This content (articles, podcasts, website info) is not investment advice. The information and opinions from me and any guests may be incorrect. The numbers and information may be wrong. The views expressed may no longer be relevant or accurate. Investing is risky. Do your own research.