FEATURED FEMMES INTERVIEW WITH HAN ZHU FROM VANKE CHINA
Interviewed on June 3, 2020
Interviewed by Yingying Zhu
Han Zhu is the head of Intelligent Data Platform for VANYI Technology, a Proptech company fully owned by VANKE, one of the largest residential and commercial developers in China with annual revenue exceeding $50 billion in 2019. She built from scratch their data analytics & BI team of over 40 people and has been with VANYI for three years now. Han was awarded one of the “30 rising stars for digital economy” at World Internet Conference 2019.
Before VANYI she worked as a senior manager in cost analytics at Walmart’s China headquarters, before which she worked at Tesla global headquarters & Walmart global headquarters in the US focusing on people analytics. Han holds Master’s degree in Statistics from UPenn and Bachelor’s degree with double majors in Mathematics & Psychology from UIUC.
What data intelligent initiatives are VANYI taking? Over the past couple of years, VANYI’s data team has basically completed Data 1.0, i.e. the digitalization of business data through applying various third party and internally developed ERP (Enterprise Resource Planning) and BI (Business Intelligence) systems and improving efficiency of the business management system. VANKE’s largest business line, based on revenue volume, is for-sale condos. We started Data 1.0 from the full life cycle of these for-sale condos projects, from design, construction to inventory, sales and property management. We designed front end interface for the sales department, so senior management can visualize the sales progress in real time. We collected and updated inventory data across hundreds of projects and made suggestions on the sale velocity. Currently, my team is working on Data 2.0, which is about data mining that can in turn allow us to gain better business insight and to capitalize on unrealized opportunities. One of the successful attempts is supplier credit analysis through establishing a database for their corporate financial and default history. We are also expanding the digitalization of business data onto our long term holding properties as well.
What was the most difficult in promoting data application in a large real estate company? The real estate industry has a low digitalization level in general. Corporate headquarters couldn’t see real time sales report across different regional offices and there was a lot of manual work in the accounting process. People often think of the cutting edge when mentioning data technologies. But our work often has to start with basic and sometimes tedious data preparation before the power of data can be actualized. For example, we started to systemize project names and the definition of project size, setting them consistent across different ERP systems. We sat together with various departments and reviewed the workflow of their business unit to find the most efficient way to incorporate digitalization. VANKE has 40+ regional offices and allows high autonomy, which is a great thing for business but makes it a challenge for my team. We must allow flexibility at the front end but still allow for data consolidation at the back end.
“The quantity and quality of data decides how much real value can be created... Through the effort of building up BIM and IoT capacity, we aim to collect high quality data in large quantity into different systems automatically.”
What is next for VANKE’s data intelligence project? The goal is “Data as asset”, i.e. having data create value for our business directly and automatically. The quantity and quality of data decides how much real value can be created. We used to input property information into ERP systems manually which could be slow and inaccurate. Through the effort of BIM (Building Information Modeling) and IoT capacity, we aim to collect high quality data in large quantity into different systems automatically. Another example is that we are now making connections between our data from the sales sector and the property management sector to better understand our clients’ purchasing preference and behavior throughout the full cycle.
How do you manage data analytic projects serving different business needs? Our data team is divided into sub teams and each sub team serves a particular business unit. The product manager of the sub team will be responsible for understanding and meeting the need of his or her assigned business sector or region. We prioritize those individual projects that has potential to become a universal platform for the entire group. I would also require the product manager to design or to procure modularized products so that they can serve for wider range of user cases while limiting customization.
How do you collaborate with VANKE’s business teams? This is the most difficult part for any IT function within a corporation. I will have to get as much support as I can, top down and bottom up. Senior management will adopt our data products as part of the performance review criteria and promote the data analytics tools that we develop to the business units. On the other hand, my team and I will take the initiative to partner with business units and address their needs by conducting as many interviews as possible. Some business units have stronger desire in digitalization to reduce manual labor or to get real time analytics to inform business decisions. We will pilot then with one or two of these regional offices that have the strongest interest. Once we have successful cases, it will be much easier to push similar data application to the rest of the group. One interesting thing is that my team has organized two VANKE data analytics competitions. We recruited those colleagues that deal with excel and other data applications on a daily basis into a training camp and taught them what our data platform can do and some basic modeling skills. But more importantly, we aimed to promote the habit of data driven decision making and hoped they would become a branch of our team to bring the positive impact to the business unit that they came from.
What new technologies does VANKE adopt, other than the data related? There is no doubt that real estate companies have to innovate in this market. VANKE is one of the most open minded and the earliest to take actions. VANKE started this digitalization effort four or five years ago. At this moment, our focus is on those solutions that can further enable technologies in our main business functions, such as smart building, energy management, robots, virtual tour, etc. In the longer term horizon, VANKE has founded a tech focused subsidiary called VANYI Technology, which will focus on developing tech products that will serve both VANKE’s internal business units as well as the whole external clients in the whole real estate industry.
What’s the difference between the status of data application in US and China? The difference is not so much about the user case or algorism. The biggest difference, according to my experience having worked in the data industry in both countries, lies in the digitalization level and the habit of making data driven decisions. I was surprised to find that a lot of Chinese companies, especially in traditional industries including real estate, tend to share their business intelligence on a one-to-one basis. In Silicon Valley, even people working for small startups are used to locating information they need from various data platforms. Therefore it is a critical part of my job to impact the business community to develop the habit of making data driven decisions.
What did you learn working for Walmart and Tesla’s Data team? I acquired data analytics skills that are obviously applicable across industries and user cases. I learned a great deal from many precious lessons especially given I joined both Walmart and Tesla’s Data team from their very beginning. Walmart, although a retailing company, faces many similar challenges to VANKE by being an offline operator and an incumbent comparing to the online retailers. Companies like Walmart and VANKE have to transport all their offline data online and go through many trial and errors in the process because technologies don’t come natural to them. For that reason, I was prepared for tough times when I first started at VANYI. Tesla, on the other hand, is a total disrupter and founded by an Internet-era star entrepreneur. Tesla adapts from the way to build software and applies it to build automobiles. They have technology in every part of their genes. They disrupted the automobile industry successfully and beat many incumbents. VANKE has to learn from the old to stay true to themselves but also from the new to stay competitive. That is why VANKE, as a real estate conglomerate, went out of their comfort zone to build VANYI Technology.
“Entrepreneurial spirit is critical from zero to one… I find great interest and become good at being a trailblazer – setting my own goals and steps, and getting things done in a way that nobody has tried before.”
What do you think are the most important skillset being the head of Data for VANYI? Over the past few years, we have walked a long way. We built the team from zero to one at the beginning. Now we have many more opportunities to make a greater impact, or what I liked to refer to as “one to ten”. The key for the two stages is different. Entrepreneurial spirit is critical from zero to one. As Elon Musk said, it is import to reason from First Principles as opposed to reason from analogies because when you try to create something new, there are not analogies and one has to go back to the roots. Many data scientists are used to certain modeling frameworks. I find great interest and become good at being a trailblazer – setting my own goals and steps, and getting things done in a way that nobody has tried before. Managerial skills that includes finding opportunities, optimizing workflow, prioritizing tasks, sourcing and consolidating resources and managing a team, becomes extremely critical in this next phase of “one to ten”.
How do you manage your own time? I spend 50%-60% of my time on cross-department communication. I travel a lot and sit down with business associates from sales, operations, regional managers and group senior leaders to understand and balance each of their needs. Data often serves as a “connector” in business. I am good at “connecting the dot” by relating all players and figuring out how their different needs can be solved in one data platform. The rest of my time is kept for my team to discuss technical topics. I would like to keep “hands on” on our data analytics work. I started from data science research myself and chose to go down the management path as this is where my interest and strength lie. But I want myself to keep rooted.
Which new technology are you most excited about? Personally I am very interested in technology’s impact on people’s cognition and wellness. I majored in psychology and math in college and always care a lot about people’s experience. In this Internet era, numerous efforts are made to optimize people’s online experience through the overwhelming social network and online shopping outlets. Real estate, on the contrary, is a natural container for offline experience. Our experience at home, workplaces and shopping places are far from being fully captured or quantified. I am very curious about a few new technologies focusing on emotional recognition and cognitive computing. I love the recent Netflix show Upload which told a futurist story about human consciousness.
Which female technology professional inspired you the most? My first boss at Walmart, Elpida Ormanidou. She built Walmart’s global people analytics teams from scratch, which reformed the business in many ways and stayed ahead on the market. She studied business analytics as opposed to data science. But she has strong belief in data driving the world and therefore fearlessly pushed Walmart’s digitalization forward no matter how challenging it could be. She is a strong leader in empowering the team to grow together with her. She can also be tough when needed:) Nevertheless, always be right-thing orientated is one of her greatest impact on me.
Translated from conversations in Chinese. Edited and condensed for clarity.