Interviewer.

Hello everyone, today we have Aijun Cui here for our CDA licensee interview! Cui Aijun can say hello to everyone!

Guest.

My name is Cui Aijun, graduated from South China University of Technology, currently working in Time China, a top 50 real estate company, as the head of the Group Data Center Department, mainly responsible for the planning, construction and operation and maintenance of data-related platforms, and also responsible for data architecture, data standards, data quality, metadata, master data and other data governance work.

 

Question 1.

It is still relatively new to do data analysis in the real estate industry, can you introduce a specific case of relevant industry data application to us?

GUEST.

Okay, I know that people have the impression that the real estate industry is relatively traditional, its management methods are relatively crude, and the digitalization process is relatively slow, but in fact, in these years, with the impact of the Internet + wave, the digital transformation of the real estate industry is in full swing, and we, Time China, are considered to be a real estate company that started the digital transformation early, from 2017 to now, after more than 5 years of construction, we We have built more than 80 sets of information and data systems or platforms, covering all business lines of the Group, enabling business data to be updated from second to day level with multiple frequencies, and enabling statistical analysis of business.

When it comes to data application cases, let's take our concern about cash flow and profit in the real estate industry. In addition to the efficient statistical summary of cash flow and profit that has occurred in the Group, we have also built a measurement model to forecast future cash flow and profit, and we have integrated business data from multiple data sources such as project master data, building product price data and room sales data, overlaid the project We integrate business data from multiple data sources such as project master data, building product price data, room sales data, etc., superimpose project node information and related sales plan, set up forecasting rules and business criteria for multiple business lines, and after calculation in the middle desk, we can easily output cash flow and profitability for the next three months, six months and one year by month, supporting our senior management to make business decisions in multiple scenarios, such as investment and financing, project expansion, sales layouts, etc.

 

Question 2.

I see you are responsible for more data governance work, what is data governance and is it different from data management/control?

GUEST.

In fact, they are sometimes really interchangeable, and their specific differences are mainly due to their different focuses.

Data governance focuses on the top-level design and strategic planning of the enterprise, and is the general outline and guidance of data management activities, which indicates what decisions to be made in the process of data management and who is responsible for them, with more emphasis on the organizational model, division of responsibilities and standard specifications, as well as the blueprint planning of data governance, the overall goal of governance, the path of governance steps and the implementation plan of governance.

Data management focuses on executing and implementing data governance strategies and giving feedback in the process, emphasizing management processes and systems, covering different management areas, such as metadata management, master data management, data standards management, data quality management, data security management, data service management, data integration and other modules of management approaches and related landing.

Data control focuses on the execution level, which is the specific implementation of the various measures involved, such as prevention, early warning and corrective measures in the process of data development such as data modeling, data extraction, data processing, data processing and data analysis. The purpose of data control is to ensure that data is managed and monitored so that data can be better utilized.

In general, data governance emphasizes top-level strategies, data management focuses on processes and mechanisms, while data control focuses on specific measures and means, and the three are complementary to each other. So many times when describing this whole piece of work, we often use data governance for generalization.

 

Issue 3.

In the face of uneven data quality and security risks, what should be done?

GUEST.

Uneven data quality is indeed a very common data problem, which is mainly manifested in incomplete data on the one hand. Part of the relevant information is not included in the system management, thus weakening the role of the application software; on the other hand, the data is inaccurate. Errors in the information provided, data errors caused by staff entry errors or system technical problems, and inconsistent data collection standards can lead to the generation of junk data, which affects the accuracy of a series of applications based on it, and makes it difficult for the data analysis system to perform its normal functions.

I think there are three main countermeasures that can be taken: First, we need to complete the online business, improve the coverage of the system, migrate the core business processes and business operations to online, and realize the online and timely update of core business data. Second, to establish standards for master data, core business data and indicator data, standardize the requirements for system entry and data development of data, implement responsible persons, rolling training, regular assessment, and strict rewards and punishments. Third, to establish the group data quality management system, in accordance with the PDCA cycle, through a combination of manual reporting and automatic monitoring to timely find data problems, organize rectification and error correction, and continuously improve the quality of data in the system. At present, these points have basically been implemented in our Time China, and in general, they are still effective.

 

Question 4.

How should the data assets of the real estate industry be built?

GUESTS.

1. Operation and management goal-driven. Can not be built for the sake of data asset construction, but to serve the current business management objectives, such as refining the management of supply, sales and inventory, improve the cost efficiency ratio, targeted asset inventory, governance and construction.

2、Governance from point to point. After drawing the complete data blueprint, the data closely linked with the operation and management objectives are governed first, closely linking business and IT, collaborative management and business frontline, and jointly completing the data standard setting, indicator setting, IT system transformation and management process building of the corresponding theme. After meeting certain goals of data analysis for operation and management, we will gradually supplement the complete blueprint in phases, and the parts that have been governed should be dynamically managed.

3、Progressively improve the asset management framework. The construction of data assets cannot be fully rolled out, but one line of service management objectives, the priority of fast-impacting work will be raised, another line of comprehensive and thorough planning, the important but not urgent business system asset inventory, data security governance, metadata management and other work gradually supplemented and complete, while the introduction of relevant tools and components of the data asset management platform.

 

Question 5.

Why does the real estate industry need a data center, and how does it work to realize the value?

GUEST.

The concept of data center was introduced to the industry by Alibaba a few years ago, and then it quickly set off the center fever, now there are different understandings and various definitions of data center in different industries, in my opinion, data center can be understood as an upgraded version of traditional data warehouse, which can contain multiple subsystems, and the biggest difference between it and data warehouse is that the high-value data it releases needs to be embedded in business operations, so the quality of data is very high. Therefore, in addition to the functions of traditional data warehouse, the data governance and development management have been greatly enhanced to strongly ensure the high data quality. Data services, which can share real-time data and models to business systems through API integration, allowing the refined high-value data to be directly embedded in business and empowering business systems. We have made the risk analysis of supplier bidding into data services and embedded it into the supply chain management system through API. When the business conducts bidding operation and selects the shortlisted suppliers, the data service returns the risk information related to the suppliers in real time, whether these suppliers have the risk of bid-rigging and whether they have conflict of interest with the company's internal staff, which greatly reduces the risk of bidding management.


Translated with www.DeepL.com/Translator (free version)