While the “996” work model seems to have become standard in the IT industry, Microsoft Japan recently launched a new experiment to allow employees to “do four breaks and three” a week on the premise of equal pay. The results of the experiment were startling: productivity increased by 40% despite the reduction in working hours. As the global IT industry’s leading enterprise, Microsoft is undergoing a comprehensive transformation from a production organization model to human resources.
If Satia is the helm of Microsoft’s transformation, then Zhao Is the practitioner of this transformation in China, from his interview, we can see that Satia in the transformation of the culture is the key, empathy, growth-oriented thinking model and other ideas have been accepted and internalized.
Zhao was Microsoft’s chief of staff for Greater China and is now General Manager of Artificial Intelligence and Digital Transformation for Microsoft Greater China. He continues to try to spread microsoft’s transformational experience, Microsoft’s concept of digital transformation, to Chinese companies.
In an office on the Beijing campus of Changjiang Business School, Mr. Zhao, who is in a suit, gave an interview to the Economic Observer, detailing how Microsoft has felt since its transformation and the blind spots and misconceptions of companies in digital transformation.
This day is also Microsoft China and Changjiang Business School to create the community and SoftBank jointly launched the “Turing Project” first phase, for industrial upgrading of industrial leaders and innovative technology as the core of the start-up enterprises, committed to AI better landing, which is Microsoft in China landing AI industry application is an important part.
During the conversation, Zhao repeatedly referred to the importance of culture, just as Satya kept emphasising the Cculture representation among CEOs and stressed that “culture remains the same and there is no hope of renewal”.
On October 24, Microsoft reported net revenue of $10.7 billion, up 21% year-on-year, for the year ended September 30, 2019. Azure Cloud revenue grew 59%, server products and cloud services revenue increased 30% year-over-year, and enterprise services revenue grew 7% year-on-year.
“Recently, we announced a broad market partnership with SAP, making Azure the first choice for every SAP customer. Our partnership with VMware and Oracle also brings these ecosystems to our cloud. We’re expanding from the cloud to the edge, enabling customers to generate data while getting real-time information while ensuring security and privacy. “GitHub has grown to more than 40 million developers, growing by more than 30 percent since it was acquired a year ago, and more than 2 million organizations are using GitHub,” Sathia said on a subsequent conference call. “
From Microsoft to customer-oriented
What many people don’t know is that in the late days of former CEO Baumer put forward the concept of “move first, cloud first”, Microsoft went from “letting every table and every family have a computer” to “let every table no longer have a computer” during the Baumer era, announcing Microsoft’s move towards the era of technology providers. After Sattia came to power, it was revised to “give strength to everyone, every organization in the world, and achieve extraordinary results”, from the original Microsoft perspective, to think from the customer’s point of view.
In terms of organizational culture, Microsoft has pursued a “coaching” model. Instead of focusing on results and numbers, managers communicate with subordinates to understand problems and help solve them. Departments are no longer separate, but more in competitive relationships, communication with each other, customer-oriented, problem-solving.
In August, Microsoft’s “do three-off four” work system in Japan showed that labor productivity rose 39.9 percent in August from a year earlier. In order to understand the need and meaning of transformation up and down the company, Satia often interacts with employees.
Economic Observer: When Sathia came to power, Microsoft underwent a full-scale transformation, in what ways, and what happened before?
Zhao: Satia’s personality is very different from Baumer’s, he has a introverted personality, but you can see that he became CEO after the transformation of Microsoft is not small. As a sales department, we report with headquarters every quarter, and now we pay more attention to the concept of “coacher”, and management is more about asking what problems we have encountered than asking about the results.
Everyone is on the same boat, your problem is my problem, but I can use what resources and angles can help you solve. For Greater China, they will pay more attention to feedback from these regions to allow Microsoft to better participate in future product development or global strategy.
In the business, in fact, Ballmer later put forward the concept of “move first, cloud first”, but at that time more from Microsoft’s point of view.
For example, the Office software we used in Apple’s iOS operating system was developed many years ago, but it didn’t open up to run on IOS because Microsoft management at the time thought office would be a big percentage of applications on the Windows platform. After Satia came to power, from the customer’s point of view to interpret the “move first, cloud first” concept, not limited to which device.
There is also a big difference in the attitude towards open source. Before that, Microsoft was very repulsive of open source, but a few years ago, Microsoft has begun to embrace open source, when we do product development in-house, even encourage programmers, before development to see if there is open source code can be used first.
Microsoft’s Asia Research Institute had a big breakthrough in AI’s computer vision in 2016, and the most important thing Microsoft did since then was to open the entire model open.
After Satya came to power, he revised the company’s manifesto and turned into how Microsoft empowered everyone to achieve something, and on this premise, you can see more competing relationships.
Apple, for example, had an iPad pro before, and before that, Microsoft had surfaces, and the two were competitors, but our office team worked with their team before the iPad Pro came out. I hope to give office customers a better experience on the iPad pro. Then iPadpro started using a touchscreen, and we helped them improve the customer experience.
In the past, we worked with customers to put the server in place and give it a manual, and it was over. But in the cloud age, we are selling no longer a product, more of a platform and tools, in this case, the need to focus on how to help customers solve problems, for their ability.
And more importantly, we are not professional experts, now working with customers, more like a joint development, co-creation of the way.
There is also a shift, which is to require a lot of continuous learning. Microsoft headquarters has a variety of courses, as well as certification, we even the President Secretary to test our cloud platform certification.
Previously, Microsoft departments were separate, and now all Microsoft employees receive an email every week. This email will help you analyze where you’ve spent the past week, it’s an interpretation through your email, meetings, messages, through AI, who you’ve worked with most, how much time you spend in a meeting, and it’s even going to advise you if you’re going to have more interaction with someone.
Since Satya came to power five years ago, he has been regularly and regularly at headquarters every month, most of the time, but he can ask questions through webcasts and on an internal platform because the goals of the transition require bosses to repeat itself.
The Blind Spot of Digital Transformation of Enterprises
“If Microsoft can’t work more deeply with Chinese customers and partners, it shouldn’t be an international company. “
Zhao remembers what he said during a meeting with Satya when he was chief of staff for Greater China.
As Satya once said, “The work of multinational corporations is more important than ever, and it needs to operate around the world, contribute to local communities in a positive way, stimulate growth, increase competitiveness, and bring opportunities to all.” Zhao said Satia has a lot of emphasis on partners, and he believes there must be a platform, an ecosystem, to work with chinese customers and partners to grow.
In the enterprise AI transformation and digital transformation, Zhao found that most of his customers have a blind spot, that is, too much emphasis on technology, and ignore what is the most suitable part of the enterprise. In addition, digital transformation requires that enterprises have sufficient data capabilities, and that communication between superiors and subordinates is enhanced by culture.
Economic Observer: You were an executive in Greater China early on.
Zhao: The most important thing in the Chinese market is that the innovation on the whole application surface, if aWe says, basic research may be a distance between China and the United States, but the most powerful thing in China is the innovation of the application surface, through these technologies to produce different applications, such as WeChat, drip truck, mobile payment and so on.
The other is that many Chinese enterprises at the beginning is the customer finally become and customer and partner, we work with them, using Microsoft global platform, let enterprises go out to sea, to Europe, Southeast Asia, Africa.
Economic Observer: In the cloud era, data security is more and more attention by enterprises, digital transformation process, many enterprises will not rest assured that the data on the public cloud, but more willing to put on the private cloud. Does Microsoft encounter such a situation, how to solve it?
Zhao: First of all, Microsoft is just a provider of platforms and tools, do not do retail stores, do not do cars, do not open a bank, we will not take your data into the retail industry to do analysis. Second, the cloud this matter has been discussed with many customers many times, I usually ask customers, if the company gave you 10 million yuan today, but cash, you take home locked in their own safe or go to the bank in the bank vault? The public cloud is our core business, and we can never afford to damage our credibility and reputation in any way, so of course we will do the best protection of these things.
Economic Observer: What are the blind spots in the process of digital transformation?
Zhao: Many enterprises have a blind spot, digital transformation is too focused on technology. When the enterprise is dominated by technology, there is a problem. The market is not really accepted, and sometimes the latest technology is not necessarily the most acceptable to the market. AI is absolutely not equal to digital transformation. The key to digital transformation is to say what to turn and why.
I’ve been to several factories before, and there’s a class of factories that I want to do so-called defect detection through computer vision. If it is engaged in chip manufacturing enterprises, the wafer unit price is high, this time to do computer vision, is worth investment and discussion.
But we’ve also been to traditional factories, such as those that make cloths, and we’ve found that artificial intelligence is really engineering. The boss thinks I only five yuan a foot cloth, but do computer visual testing, because the pixels to be high enough, the camera even to hundreds of thousands or even millions of yuan, the boss thought that, since this is the case, it would be better to ask two more people with the naked eye to see it.
If there is no before these business pain points and culture have a certain understanding, technology alone to do, may not be the most suitable for customers, in fact, like the traditional cloth factory such customers, but do not need to use computer vision to do defect detection, with our most traditional big data analysis can actually do almost the function.
Another problem is the amount of data capacity. Researchers understand AI, very understanding of technical learning models, deep learning models, can solve the model very well, but the model is not the customer wants, the customer wants you to help me reduce the defects of my cloth, you help me reduce the problem of shipping.
If not ready, there is enough data quality and quantity, as well as enough data capacity, directly and do digital transformation of the enterprise to talk about cooperation, but will finally let the boss feel that AI is useless, then forget, stop this side, it is a pity.
A lot of times is not AI useless, more is there is no around the matching can make AI use well.
The other is communication at the top and bottom, because the most understanding of the problem is the front-line staff, but few companies can give front-line employees enough space to speak out. One of our partners, spent half a year to solve a problem, very excited to go to the production line landing, the production line of the master said, this problem we have solved, is by virtue of engineering experience, but the boss does not know, or his supervisor does not know.
Another important thing is that the data will change, Microsoft had a very painful experience before, we and Seattle High School with big data analytics, hoping to find out which students need additional remedial or strengthening, very good at first, the overall school performance improvement is very high. A year later, the school stopped the whole project. Why? Because student data is changing, last year’s looks good data models may run into problems in the second or even third year.
Many foreign companies now say the Chinese market is the best place to train because Its customers and competitors are getting too fast. However, when developing some data models for the Chinese market, it is important to emphasize that the model should be continuously optimized.
“I firmly believe that we need not only a new wave of technology, but also a new wave of personal information protection, a new wave of security, and a new wave of ethical and personal rights issues that prevent artificial intelligence from occurring. Brad Smith, President of Microsoft, said at the Web Summit Summit in Lisbon on November 6. The terms “policy” and “technology” have been mentioned many times at Europe’s largest technology summit.
Economic Observer: In the face of the consequences and impact of new technology, enterprises should bear what kind of responsibility?
Zhao: When AI started to help us make decisions, we designed these AI software, in fact, we had a lot of responsibility. Responsibility AI in human-machine interaction is a very important element, you need to give users a feedback mechanism, otherwise you originally want to attract more customers, but through this customer more disgusted.
Throughout the user experience, you must make sure that you have to let the customer know exactly what you’re doing with AI, and even put the accuracy of the AI on top of the data.
In addition to this, there is a very important piece of data bias. For example, face recognition is the most mature in AI technology, because we have a lot of data, any Hollywood movie, any film and video, photo is all face recognition one of the data, claiming accuracy can reach 99% or higher.
But what does 99% equal to what? Last year, a New York Times reporter did an analysis that if you’re a white male, you’re 99 percent accurate. If you have dark skin and a very dark sun and a woman, your accuracy is only about 70%.
It’s not that programmers are racist, it’s that the data itself is biased, but they didn’t know it at the time of development. But the impact in the middle is huge. Microsoft has created a mechanism to let users know which is wrong.