Author: Wu Nü Wang (Peng Zhao)
IoT Think Tank Original
Please indicate the source and origin when reprinting
Introduction
This week, I had an in-depth conversation with Mr. Liu Kezhen, Chairman of Advantech Technology. Based on the experience accumulated from long-term industry immersion, he pointed out certain “emperor’s new clothes” phenomena—such as the absence of business models in IoT platforms at this stage. For the currently popular data centers, he has a similar judgment, that it is also difficult to find a business model. Therefore, in this article, we will think together: 1. Why has the data center rapidly gained popularity? 2. Why do we say that IoT platforms and data centers lack business models? 3. Where is the way out for IoT platforms and data centers?
This is the 133rd article I have written in the “Wu Nü Heart Sutra” column.
This week, I had an in-depth conversation with Mr. Liu Kezhen (Mr. KC), Chairman of Advantech Technology.
Founded 36 years ago, Advantech Technology has grown into a multinational company with annual sales exceeding $1.5 billion in industrial automation products and solutions, and is now advancing towards IoT and edge intelligence.
Based on the experience accumulated from long-term industry immersion, he pointed out certain “emperor’s new clothes” phenomena.
For example, there is currently no business model for IoT platforms.
“Although Advantech has its own WISE-PaaS IoT platform, I believe that relying solely on the IoT platform cannot generate profits; the platform is meant to carry services, not to profit.”
For the currently popular data centers, he also has a similar judgment, that it is also difficult to find a business model.
After all, Advantech Technology is a company with a strong hardware gene, and Mr. KC’s judgment on the IoT platform may not necessarily represent the future. However, understanding and extending his logical thinking helps us further see the development of IoT platforms and data centers.
Therefore, in this article, we will think together:
-
Why has the data center rapidly gained popularity?
-
Why do we say that IoT platforms and data centers lack business models?
-
Where is the way out for IoT platforms and data centers?
01
Can the ancestor of the center still make a living?
Let’s quickly sort out some related knowledge about the center.
What is a center?
According to the 10-character definition given by Wang Jian, Chief Consultant of the data consulting company Thoughtworks, the center is:
“An enterprise-level capability reuse platform”
-
“Enterprise-level” defines the scope of the center, distinguishing it from single-system services and microservices.
-
“Capability” specifies the main objects carried by the center, and the abstraction of capabilities explains the existence of various centers.
-
“Reuse” defines the core value of the center; past platformization did not pay enough attention to reusability. The rise of the center has shifted people’s focus more from the platform’s internal aspects to its support for front-end business.
-
“Platform” indicates the main form of the center, which is different from the piecemeal approach of application systems. By recognizing and platformizing finer-grained capabilities, it achieves flexible reuse of enterprise capabilities and better supports front-end business.
Who is the ancestor of the center?
The concept of the center was first proposed by Alibaba in its “big center, small front” strategy in 2015. Its inspiration comes from the small Finnish company Supercell, which has only 300 employees but has continuously launched blockbuster games, making it the world’s most profitable star game company.
In mid-2015, Jack Ma led Alibaba’s executives to visit Supercell. In June 2016, Tencent announced the acquisition of 84.3% of Supercell’s shares for $8.6 billion.
What is the magic of Supercell? It is precisely this small company that pioneered the “play style” of the center and applied it to the extreme.
This seemingly small company has set up a powerful center to support numerous small teams in game development. This way, each team can focus on innovation without worrying about the foundational yet crucial technical support issues.
Supercell’s CEO, Ilkka Paananen, manages a game company like a professional sports team. He believes that the only mission of management is to attract the best talent, create the best environment for them, give them freedom and trust, help them overcome difficulties, and make the company a place where the best talent can have the greatest impact.
Everything else, including financial goals, is secondary.
Thus, Supercell has constructed a completely inverted management structure.
The traditional management structure is pyramid-shaped, with the CEO often at the top. The greatest innovation of Supercell lies in its management structure, which is completely inverted. Paananen’s proudest label is: “the CEO with the least power in the industry.”
Supercell’s overall structure adopts a “developer-led” model. The 300-person team is divided into several small teams, with 5-7 game developers forming a small team, developing their own games and launching public beta versions at the fastest speed to test the popularity of the games among users.
These small teams are also known as “cells”; Supercell is the collection of these cells, which is also the origin of Supercell’s name.
Thus, it can be seen that the center is not merely a system or platform, but also a reorganization and transformation of the organizational structure.
However, the business process of enterprises is influenced by various factors. The center is not a panacea; it has not allowed Supercell’s sales to maintain a rapid growth for long.
The reversal of the situation had already emerged in 2016, and in 2017 and 2018, Supercell’s revenue experienced a significant decline for two consecutive years.
What types of centers are there?
According to current common statements, there are six types of centers:
-
Data Center: Provides data analysis capabilities to help enterprises learn and improve from data, adjusting direction.
-
Business Center: Provides reusable services, such as user centers, order centers, and other out-of-the-box reusable capabilities.
-
Algorithm Center: Provides algorithm capabilities to help offer more personalized services and enhance user experience.
-
Technology Center: Provides technical support capabilities for self-built systems, helping to solve underlying technical issues such as infrastructure and distributed databases.
-
R&D Center: Provides management and technical practice support capabilities for self-built systems, helping to quickly build projects, manage progress, test, and ensure continuous integration and delivery.
-
Organizational Center: Provides support for investment management, risk management, resource scheduling, etc.
Due to the close relationship between data centers and the Internet of Things, some companies have already integrated data centers as part of IoT platforms to provide services externally.
Therefore, I will focus on data centers as a typical representative.
How popular are centers?
The term “center” began to enter people’s vision on May 21 of this year.
On that day, Tencent held a global digital ecosystem conference, where Tencent executives proposed: “Open center capabilities to assist industrial upgrades.” Since then, discussions around centers have begun.
Some even refer to 2019 as the “year of the data center,” highlighting its popularity.
The rapid rise does not exclude the possibility of over-packaging. Some netizens commented: “Internet companies love to hype concepts. What was originally very clear becomes profound after their packaging.”
Why has the data center rapidly gained popularity?
Regarding this issue, I agree with the viewpoint of Wang Xin, a senior partner at Roland Berger Management Consulting: The data center is merely the surface waves stirred up by the sea; the underlying undercurrents are changes in the commercial environment.
He believes: “In the context of rapid business growth, no one cares about how the organization is built, because they are too busy seizing market opportunities; however, when the environment is not so optimistic, internal management will turn towards lean.”
The economic environment is tightening, and both internet and traditional enterprises face transformation pressures. However, under such demands, many companies have realized that in the era of rapid business growth, the side effects of solely pursuing financial growth have begun to emerge, and the “walls” between departments within the organization have become deeply entrenched.
Based on traditional organizational structures, the assessment and incentive mechanisms across departments differ, and collaboration faces numerous obstacles. Ordinary employees’ bottom-up innovation is nearly impossible.
In the face of the current situation, enterprises need to modularize and reuse various capabilities through the center to better serve industrial customers. Many enterprises have realized the enormous value of data and hope to utilize it well; thus, the demand for data centers has emerged.
02
Why do we say that IoT platforms and data centers lack business models?
The center is a reorganization of organizational structure and a new thinking system.
As the center focuses more on the reuse of capabilities, data centers add flexibility to IoT platforms, making it easier for data to drive value creation.
The center reconstructs the originally “siloed” application systems into easily callable “cells,” achieving the reusability of capabilities. This system design philosophy is not a new thing. By utilizing center thinking, when adding or modifying functions on the platform, the workload and difficulty are greatly reduced, and collaboration between different teams and departments becomes easier.
A good platform generally needs to first become a good center.
In IoT platforms, there are more or less shadows of data centers. However, achieving this requires years of technical accumulation and business sedimentation; it cannot be accomplished overnight.
At the same time, both IoT platforms and data centers face a common challenge: how to find a business model?
Let’s start with IoT platforms.
Why does Mr. KC believe that IoT platforms lack business models?
His viewpoint is that the profitability of IoT platforms must come from outside the platform.
IoT platforms tend to be infrastructural, service-oriented, and entry-oriented, much like search engines on the internet or operating systems on computers or mobile devices; they cannot make money just by being a platform. The platform profits indirectly from the hardware it carries or from the SaaS services provided on the platform.
Here, it is necessary to clarify a misconception: Mr. KC mentioned that he encountered some companies that clearly only did an IoT project or built an IoT application but insisted they were IoT platforms. These “platforms” are not within the scope of his discussion.
Currently, the profitability basis of IoT platforms is still very weak. Only when the SaaS “co-creation” industrial ecosystem on the IoT platform is very prosperous can the platform possibly profit.
In other words, the value of IoT platforms needs to be reflected through ecosystems.
To accelerate the commercial returns of IoT platforms, it is essential to prioritize the prosperity of the co-creation ecosystem at the SaaS layer. From the practices in industries such as industrial, medical, retail, and building, Mr. KC believes there are four forces that contribute to the co-creation ecosystem in SaaS:
-
The first type is the IT teams within large enterprises. For example, large hospitals may have more than a hundred IT engineers, belonging to the in-house co-creation SaaS ecosystem.
-
The second type is traditional system integrators transforming into new IoT integrators while also providing SaaS services.
-
The third type is startup IoT service providers, many of whom are ISVs (Independent Software Vendors) with native IoT genes, directly providing SaaS services to customers.
-
The fourth type emerges after the first type matures, becoming newly derived IoT SaaS service providers that empower externally.
According to the evolution process of digital transformation, each industry generally goes through the following spiral ascent cycle:
Visualization of existing data → After visualization, decision-makers realize the value of data, further enhancing and improving data collection capabilities → More data continues to be generated, establishing a certain data analysis foundation → New algorithms such as artificial intelligence have a place to apply, and decision-making processes continue to improve → This cycle repeats, enhancing the positive cycle of continuously exploring data value.
For different industries, the maturity of ecosystems varies, and the pace of digital transformation in industries differs. Many industries have few and small partners willing to co-create ecosystems with IoT platforms, making the commercialization path of IoT platforms quite challenging.
Based on the above considerations, Mr. KC divides Advantech’s IoT platform and co-creation ecosystem development path into three stages:
-
Stage 1: Expand the existing embedded hardware platform. This is the foundational part of data collection, such as edge computing and terminal products.
-
Stage 2: Build the WISE-PaaS industrial IoT platform centered around industrial PaaS. Currently, WISE-PaaS has launched multiple integrated application solutions (SRP), including Overall Equipment Efficiency (OEE), Energy Management Systems (EMS), Fault Prediction and Health Management (PHM), and many other modular units.
-
The ultimate goal of Stage 3 is to achieve IoT application integration platform solutions. Through cooperation with vertical industry partners, the IoT solutions will eventually be rapidly popularized in industries such as industrial, energy, environment, and smart cities.
Now let’s talk about data centers.
The value of data centers is reflected through integration.
In the past era of rapid economic development, many companies inevitably encountered the phenomenon of “reinventing the wheel” during large-scale business expansions.
If the center wants to perfectly solve the “reinventing the wheel” problem, it requires not only technological iteration but also organizational change. Therefore, only when center thinking is integrated and embedded within the enterprise can practical efficacy occur. Thus, the implementation of data centers must be analyzed on a case-by-case basis, often driven by the dual services of technical implementation and project consulting.
Similar to the IoT platform ecosystem, there are very few enterprises capable of implementing data centers. Compared to ISVs, finding suitable IDVs (Independent Data Vendors) is even more challenging.
Even if found, it may fall into another misconception.
Professor Li Jie, founding director of the Intelligent Maintenance Systems Research Center (IMS) of the National Science Foundation (NSF), mentioned in his new book “Industrial Artificial Intelligence”:
“We have encountered many data service companies interested in industrial scenarios, and they all share a common characteristic: they first ask us what data we have, rather than asking what problems need to be solved.”
He believes this phenomenon is caused by differing perspectives between both parties.
Whether to start from the perspective of data to find hidden relationships and application opportunities, or to start from problems to create value, will lead to entirely different results.
03
Where is the way out for IoT platforms and data centers?
The year 2019 marks the integration and reshuffling stage for IoT platforms, and even large enterprises cannot escape.
In March of this year, SAP laid off all top HANA developers. HANA was originally a business data platform, but industry insiders claimed that HANA exaggerated its technological innovation. Recently, insiders revealed that SAP’s IoT brand Leonardo has become a mere name, and the development team for SAP’s IoT PaaS platform has also been disbanded.
In previous articles, I mentioned that there were once over 400 IoT platforms globally, but the development paths that can survive are now relatively clear, generally falling into two categories:
Path 1—Horizontal Development: Follow the horizontal cross-border platform route, build an IoT infrastructure-type platform, do not prioritize short-term profits, and continuously expand market share.
Path 2—Vertical Deep Cultivation: Deeply cultivate a specific vertical industry, establish a moat by solving industry pain points, develop vertically, and seek to bind with leading enterprises in the industry.
Enterprises on these two paths each have their own characteristics:
Path 1 is often favored by giants, with “behemoths” from the internet, telecom, IT companies, and traditional enterprises scrambling to occupy their respective territories.
The IoT platform is an area with a Matthew effect, where the leading IoT platforms will be very few.
The final outcome remains uncertain, and given the unique development rhythm and objective laws in the IoT field, even large entities like GE, IBM, and SAP may not necessarily benefit.
Path 2 is often chosen by innovative IoT platform companies. After all, with limited resources and funding, it is better to focus on a single industry and continue to deepen. For example, in China, a batch of smart home IoT platform companies started running around 2012. Around 2015, another batch of industrial IoT platform companies gradually emerged.
Similarly, there are many IoT platforms abroad that focus on a specific industry. Compared to the economic returns obtained through direct sales from the platform, the collaborative and amplifying effects these platforms create for traditional industries have more practical value.
According to a recent research report released by American investment bank KBCM Technology Group: The best exit strategy for most IoT startups is not IPO, but acquisition.
Buyers are usually large traditional enterprises with transformation or integration needs.
Similar trends have also emerged in China.
Those innovative enterprises that were the first to run in the smart home IoT platform track have successively received investment from traditional industries, especially from “golden owners” in the real estate and related fields who urgently need transformation.
-
Ouyibo received a C-round financing of 130 million yuan from Midea Real Estate and Red Star Macalline in May 2019.
-
Tuya Smart signed a strategic investment agreement with Greenland Group in June 2019, with Greenland becoming an important shareholder.
-
Yunzhiyi was invested by E-House China from the very beginning, launching an IoT cloud platform with resources from E-House Group.
-
More investment cases will soon be revealed…
Looking back, there are also many manufacturing “gold owners” waiting to take over the industrial IoT platforms.
There was once a statistic showing that companies in China with a net profit exceeding $1 billion are mostly manufacturing enterprises, including typical examples such as Weiqiao Group (textiles, aluminum, electricity, etc.), WH Group (food industry), and Conch Group (cement industry).
Manufacturing enterprises facing transformation pressure may trigger a new wave of investment and acquisition.
With the experience of IoT platforms ahead, the exit strategy for data centers is self-evident.
Finally, I would like to announce that my video conversation with Mr. KC of Advantech Technology will be broadcast soon, please stay tuned.
Summary of this article:
1. The center is not merely a system or platform, but also a reorganization and transformation of the organizational structure.
2. The rise of data centers is merely the surface waves stirred up by the sea; the underlying undercurrents are changes in the commercial environment.
3. The value of IoT platforms is reflected through ecosystems, while the value of data centers is reflected through integration.
4. The development paths for IoT platforms are twofold: horizontal development or vertical deep cultivation. Being acquired is the optimal exit strategy for IoT platforms.
5. The implementation of data centers must be analyzed on a case-by-case basis, and its ultimate exit strategy is even more singular than that of IoT platforms.
References:
1. Professor Li Jie, “Industrial Artificial Intelligence”
2. KBCM Technology Group, The Realities of THE EXIT
3. SAP HANA Mass Layoffs: What Happened?
4. The “Center Wars” of Internet Giants
5. Roland Berger: The Center, the “Accelerator” of Successful Agile Organizations
6. Data Centers and Industrial Internet Platforms
7. The Next BAT, the Next TMD
8. Data Centers, the Next Platform-Type Entrepreneurship Opportunity
9. The Importance of a Strong Center [Produced by Jingwei]
10. Finally, Someone Has Made the Center Clear
“Zhihui • 2019 AIoT Industry Leaders Summit”
Coming Soon
Organizer: IoT Think Tank
Date: August 8, 2019, 9:00-17:00
Location: No. 9 Jiu Xian Qiao North Road, Chaoyang District, Beijing, Hengtong International Innovation Park C9
Agenda as follows:
Gathering of Experts
Scan or “Read the Original” to register for free!!!
Business Contact: Wanwan 18910797375 (WeChat)
Summit Q&A: Leilei 18514298212 (WeChat)
