The very useful first steps would always be building mutual trust and align goals. Instead of jumping to the Hackathon topic, the team started with some introductions, chit-chats, members’ self-introductions (background, current career/academic focus, professional expertise, things to do in the city, etc). Members also went on to align the goals: the Hackathon event is more than a school project and the team would like to devote as much energy as possible to deliver it.

In retrospection, the author finds the beginning discussions very help. It motivates the team towards the same goal and helps members to understand each other’s expertise. Chinese members have purchasing experience on JD.com before and know better about JD’s image, strategies and innovation in various areas. The US members have strong tech background and contribute the most regarding data crunching, solution construction and SQL analysis.

However, the team did run into some challenges when deciding the key issue to solve in the project. The US members suggested JD to learn from Amazon while the Chinese members believed given JD’s self-operated nature and the country’s infrastructure, some sample patterns cannot be installed directly. It was until 6 in the afternoon that the team reached a consensus to utilize machine learn to find out consumers’ unmet needs and high margin areas in which JD should cultivate its own private brand (white label).