I just finished my first trip to Baltimore and came back to Atlanta. Being nominated to attend and represent Emory University in the nation’s largest Analytics event: 2018 Informs Business Analytics Conference, I had the chance to talk with management teams from the company sponsors and to learn the Analytics trend within the industry. I also participated in the Early Career Professionals’ Network (ECPN), a program inviting 20 Master or Ph.D. students majoring in Analytics from American top universities including MIT, Northwestern, Emory, Georgia Tech and USC. So I also took the opportunity to learn from my peers as well. It was an amazing experience, and here are the three things I learned from the discussions:

1. The market still has a shortage of people who have both Data Science skills and industry expertise. Traditional companies building up their own Data Science teams find it hard to acquire talents who have the combination of the both because Data Science is a relatively new area of college educations. The director of one of the largest automaker’s Machine Learning team told me that his team members are having difficulties communicating with business experts within the company. So although they have their own Data Science team, they still outsource a lot of projects to consulting firms.

2. There is a gap between traditional Quantitative Analysis and modern Data Science. Some companies choose to separate their Operation Research teams and Machine Learning teams since they normally consist of people from two different backgrounds. But the problem is that even if companies put the Operation Research people together with the Machine Learning people, they still find the two groups of people have trouble working efficiently with each other. But since Machine Learning (Predictive Analytics) and Operation Research (Descriptive Analytics and Perspective Analytics) are both needed in most Data Science projects, there is a gap to be bridged.

3. Instead of blindly following the Data Science trend, companies now want Data Science to be used to solve business problems and create business value. We discussed “How do you evaluate your work” during ECPN. The idea I provided for my team was “High Quality and High Impact”, meaning that a project should not only have high quality but should also be highly impactful. A Data Science team form a major consumer credit reporting agency in the U.S. consisted of a group of advanced researchers, and they spent the whole last year doing research and drafting a report. Since the techniques they implemented was not used before, they assumed their report would be in high demand. However, even three months after the launch, the sales were zero. The whole team ended up being fired this year. As Ph.D. researchers, no doubt that their work had high quality. It was not being able to know what is needed in the marketplace and not being able to provide the impact that caused their failure.

To conclude, I really appreciate the chance to attend the conference and become an ECPN alumni. I did learn a lot about the industry trend from professional speakers as well as from my peers, and I really look forward to volunteering as a mentor in the future and being able to keep developing the ECPN program.