Big Data Realization by Scenario-driven

Xuewu (Shawn) Wang, Head of Data Labs, China Eastern Airlines

Xuewu (Shawn) Wang, Head of Data Labs, China Eastern Airlines

Due to the development of the Internet and the revolution of innovative technologies, the market competition in various industries has become increasingly fierce. The traditional companies need to undergo data-transformation by applying big data, machine learning, artificial intelligence, NLP and other algorithms to their business operation and market development to improve their competitiveness. The MU Data Labs led by Xuewu Wang won the 2018 IDC (International Data Corporation) Digital Transformation Award: The Leader of Information and Data Transformation in China. He will share his experiences in the data transformation.

The Challenges Faced by Traditional Companies

The data era has arrived, the market, the customer demand, and the business model has changed a lot. In the past few years, the aviation industry, China Eastern has been facing increasingly fierce competition in the travel market due to the rise of oil prices, the continuous expansion of high-speed rail lines, the development of low-cost airlines, and the expansion of online travel agencies. China Eastern needs data transformation to improve market competitiveness.

Limitations Encountered by Most Companies When Using Data

There are many limitations to companies using data. The most common problems are computing capabilities of massive data, data analysis capabilities, data quality, data standards, and data islands. The biggest difference between the traditional companies and the internet companies is that internet companies can digitize the behavior of all online customers. The traditional companies have limited ability to collect the data of customer behavior. Although the traditional companies lack of service contact points that can digitize customer behavior, they still have amount of sales data, ERP data, CRM data, sensor data, and customer feedback data. However, in most cases, this data is not connected in the case of data islands. The traditional companies should establish data platforms or mechanisms to change this situation, allowing data to be shared within the company and allowing data to flow between different business fields. Building data warehouses, domain data marts, big data platforms, and data sandboxes are good solutions to this challenge. Meanwhile the traditional companies should increase service contact points, generate and collect more data and fully exploit its value.

Three KEY capabilities for Data Realization

“Although the data has not yet been included in the balance sheet of the company, this is only a matter of time.” Viktor Mayer-Schönberger said.

The transformation of data has three core drivers: data-ization capability, big data analysis capability, and data product capability.

Data-ization Capability

Datai-zation capability means the company can collect the data of all business processes and digitize business activities such as production, operation, customer service, and marketing as much as possible. Data has become a very core asset for every company.

China Eastern has 650 aircrafts, covering 177 countries, serving more than 100 million passengers per year, and transporting 1074 destinations. It ranks seventh in the world. We build business analysis systems and operational support systems for passengers before, during. and after-flight. They are digitizing the entire process of passenger travel and implementing a closed loop of passenger service. In customer services, marketing, business development, cabin services, ground services, flight transportation, aircraft engines, call centers, operations centers, e-commerce, APPs, etc., generate large amounts of data every day. We face the challenge of managing large amounts of data.

After collecting massive amounts of data, the company needs to manage the data. It would be a successful solution if the company builds an Enterprise Master Data Management system and a Metadata Management system. At the same time, to ensure the data quality and the data standards, the company needs to implement data governance.

Big Data Analysis Capability

Big data analysis capability is a core competency of a company. Efficient big data analysis capability can quickly capture the insights from massive amounts of data, discover patterns and predict the future. Companies only have large amounts of data, but do not have the ability to analyze the data. It is impossible to realize the value of data assets. Analysts need computing power that can handle massive amounts of data. They also need to have the ability to translate business problems into data analysis problems. Developing a data mindset in the process of constantly using data is very important to them. The best big data analysis practices have following characteristics: (Re) use data, data-driven mindset, continuous exploration process, iterative open & share, and verification. To implement effective big data analysis, analysts need to establish a data analysis process, and continually and iteratively optimize the analysis results so that the desired benefits can be achieved by analyzing the results.

Data Product Capability

Business objectives can be clearly defined by alliance data teams. Defining what problems can be solved by the analysis results, where the analysis results will be applied and how they are applied can help quantify the value of the analysis results as much as possible. Companies need to manage data generation, data collection, data integration, data analysis, and data productization.

There are many ways to realize the value of data, such as sales data, data products, data services and data sharing. Business experts, analysts, and data experts form an alliance data analysis team to clarify business scenarios, required resources, business goals, and implementation plans. The alliance team is driven by the business scenario to design algorithm solutions and quantify the business value of the results. They can apply data exploration, data discovery, data visualization, data mining, machine learning, artificial intelligence, NLP or operations research to implement business scenarios and quantify the value of their results. For example, the company can achieve increased profits, improved business efficiency, reduced operating costs, new market opportunities, and risk management through algorithms.

The strategic position of data in the company is becoming more and more obvious. Whoever has more data, it can understand the market better, know the situation of competitors and grasp the initiative of the market. Many companies are improving their competitiveness through big data and AI. Big data realization by scenario-driven is a very successful best practice. It will determine the right research direction for us and greatly reduce the risk of investment for the company.