Meet Matthew, Ian, and Saizy — Coupang Data team

Coupang makes business decisions based on data. To this end, Coupang has data experts designing Coupang's unique data systems and tools and has more professional experience using its own A/B system. 

We met Data Engineer (DE) Matthew, Data Scientist (DS) Ian, and Data Analyst (DA) Saizy, who connect Coupang's business with customers through data. How are the three data positions collaborating within the Coupang’s data ecosystem? 

Hello, thank you for having the interview. Would you please introduce yourself?

Matthew: Hello, I’m Matthew leading the Data Engineering team of the Data Platform organization. The team collects and processes data that affects Coupang's key business decisions and builds and operates big data infrastructures. 

Ian: Hello, I’m Ian, a data scientist from Coupang Pay. My job is to develop a credit rating model for customers and to create an engineering environment where data scientists can work in better working conditions. 

Saizy: Hello, I'm Data Analyst Saizy. I'm working on feature analysis related to customer experience. More simply, I analyze and provide data throughout the process of developing features to address customer pain points to ensure they are valid and efficient. 

You are all in different teams, please tell us more about your team. 

Matthew: The Data Platform organization is mainly divided into two teams; one working on ETL operations and data mart management based on massive internal data, and the other team creating infrastructures for various users, including data analysts, data scientists, and data engineers to analyze/process/schedule/visualize their data. 

Ian: Our team takes care of making predictions on a variety of products related to Coupang’s different payment products. The product we are currently focusing on is PayLater, which is a BNPL (Buy Now Pay Later) service. Our job is to lower the delinquency rate by building Coupang’s own credit rating model based on data.    

Saizy: The DA releases several features to improve Coupang's customer experience. As an analysis team, we help decision makers make data-driven decisions before and after the release of features. 

Which Leadership Principle best describes your team’s culture?

Matthew: It's ‘Company-wide Perspective’. As I said, the Data Platform organization builds data platforms and big data infrastructures. We operate and improve systems by considering data analysts, data scientists, and data engineers as much as possible. Also, we do our work completely from the company-wide perspective. In other words, we design and build data and systems considering not only the needs of specific business domains but also company-wide impact and internal security. 

Ian: It’s “Learn Voraciously” for our team. We develop great teamwork based on “learning.” Working with other teams, I learned and used a range of new modelling techniques which helped me to grow faster than ever before. We share a lot of insights with each other while discussing how to develop models in a way that can be well reflected in business.   

Saizy: It’s ‘Wow the Customer’. Our organization focuses most on improving customer experience. We take a deep approach to various areas that affect our customers from the Coupang app home to the page where the order is completed. We are also testing various hypotheses and improving our customer experience in a strategic and logical way. 

Coupang is a data-driven company, but what is the actual working method like? Please explain with real-world examples how Coupang uses data to improve customer experience and create innovation.

Matthew: Data marts and source data we provide have a significant impact on data-driven decisions. We review data points of key services and data marts at weekly Business Review meetings. Through the process, we analyze customer and system pain points and resource investments and make various decisions. 

Ian: Coupang has a well-established system where you look through data and make a quick decision. This system helped us to lower the delinquency rate in a short time after launching PayLater. All steps, from collecting data to building and applying an improved model based on the data, need to be taken without delay to quickly improve customer experience. And Coupang has the best working environment for all these three steps. 

Saizy: Every decision made in the organization is based on data. Tests are required to release features to improve customer experience, and data-driven decisions are made across the process. Data is essential in every decision to determine if experimental parameters are sufficient, hypotheses are reasonable, effectiveness measurements at run time are accurate, and the results are statistically meaningful. 

Data Engineer, Data Scientists, Data Analyst must work together. How do the three jobs work together?

Matthew: No matter how large the data volume is, a well-organized dataset is a must. DE secures and refines data in the process of data collection-loading-processing (ETL) on big data platforms to help DA and DS analyze data efficiently. We also provide tools for internal users including DA and DS to directly access and control data. 

Ian: You can have more fun when working with data engineers who well understand our business. Data engineers address the limitations that hinder the development of increasingly complex modeling techniques while data scientists find ways to make data useful to our business. The more we help each other, the more accurate predictions we can make based on data. 

Saizy: When the DE team provides the right environment for analysis and provides source data and data marts, the DA team actively utilizes the data and creates the necessary marts with the source data. We’re always grateful that the DE team is constantly improving the environment for better analysis. 

What was your most memorable experience working at Coupang?

Matthew: When we were working on the AWS migration project, we had a special experience of designing a new data infrastructure in the cloud. We were able to understand many businesses' use cases and how big data platforms are utilized and production pipelines are implemented. We were also able to acquire a lot of know-how about the cloud environment in a short time, contributing to strengthening the teamwork. 

Saizy: I'd like to talk about the A/B test tool. It’s Coupang’s own tool to see experimental results briefly and help make statistically reasonable judgments. While I was constantly using it, I was impressed by its convenience, scalability, and mechanism, which is never inferior to other analysis tools on the market. 

Ian: I’d like to talk about Coupang’s free but structured working environment. Before I joined Coupang, I used to work independently in a rather rigid way. But in Coupang, you have more chances to share thoughts with coworkers compared to other workplaces. 

Coupang has an environment and system that can analyze huge data from various angles. We are looking for those who want to improve customer experience by analyzing vast amounts of data in depth and setting business strategies and directions together.