Testing the Performance of a Pricing Optimization Model — my practicum journey with Engage3

Beverly Huang
Apr 8, 2019 · 4 min read

It has been more than six months since we started the project. I still remember how stressful I was at the beginning of this project.

When A Business Student Encounters Complex Algorithms

Before coming to UC Davis, I was a business student studying global supply chain management. During my internship and project experience, I realized that the future direction of supply chain management will be precise management closely related to data analytics. Thus, I decided to come to the Master of Business Analytics program at UC Davis, a program located in Silicon Valley, to enhance my analytics skills by working with companies on the forefront of the big data revolution.

I have been working with Engage3 for the past six months. Engage3 is a start-up helping retail companies make better pricing and promotion decisions by leveraging big data, predictive analytics, and SaaS solutions. The CEO of Engage3 invented a new pricing optimization model, and our team is working on testing and benchmarking the accuracy and robustness of this model.

This is a pilot project supervised directly by the CEO. At the kick-off meeting, when the CEO introduced the project, I was totally shocked by the complex machine learning models and got lost at those statistics formula and concepts. At that time, I had no idea how my life with this project would be and doubted whether I could survive from this project.

It IS A Challenging Journey

The later engagement with the company proved my concerns. It is a challenging project for me.

Our client has a well-planned and tight timeline for us. We have weekly meetings with our main contact Sahar (the firm’s lead data scientist) and she assigns weekly tasks for us. We need to closely work together to deliver on time while taking six courses at the same time.

The past 6 months was hard but fulfilling. Our main task was to design an algorithm to divide stores into the control and treatment groups to prepare for the testing experiment. To achieve that, we used historical sales data from one of Engage3’s clients to simulate demand, predict sales, cluster similar stores together, and then put them into the control treatment groups to make sure both groups are very similar so that we can start the experiment at the later stage.

Our client has a high standard on our deliverables. First, we need to be familiar with advanced knowledge of statistics and machine learning and proficiency in using SQL and Python. Second, we need to validate our codes to make sure they are readable and reusable. To achieve that, we validated our algorithm using data of another client of Engage3 to see whether our algorithm works on a different dataset. Third, we need to document our codes. We not only put docstring (i.e. explanations of what each function does) within each function we created but also wrote manuals for each notebook we created.

Improving Along the Journey

Our team is improving along the way. So am I. First, I am developing my business acumen through the journey. Not only did I have a deeper understanding of the retail pricing industry, more importantly, but I also got a taste of what working with Silicon Valley company would be like and how an innovative idea is born and tested before implementation.

Second, my technical skill set is improving. I am transferring from a pure business student to a business analyst with a quantitative mindset and skills. I now know how to identify a business problem with data and solve the problem with analytics.

Last but not least, my communication improves a lot. Through the project, I have been communicating with our clients and mentors and presenting our project status to the class and even to the company. My most unforgettable experience was our presentation to the Engage3 team at their headquarter in Davis.

Working in a “start-up” environment with my teammates

Before the presentation, we had a four-day, rigorous sprint at Engage3. During the four days, we worked closely with Sahar to fix thousands of lines of complicated codes to make everything work together. It was the most stressful period for our team. Chunks of code that were sure to run broke down whimsically. We supported each other technically and more important mentally. Finally, we conquered the tasks.

On the final day, we presented our project to the Engage3 team. This was my first presentation to a group of experts and C-suites of a company in a professional setting. At that time, I felt that I was not just a student, I was a consultant capable of solving a real business problem and communicating the story to experts in the field.

Our team and client leads after the presentation

Ready for Upcoming Challenges

This project puts my business knowledge and quantitative skills into practice, transforming me from a business student into a business analyst for a company and very likely an expert in retail pricing.

We are now just half-way through the process. We will start the experiment and monitor the performance of our model with Engage3 in the following months. We are now more confident and ready to achieve the ultimate goals of the project. We are also ready to tackle real-world problems on the way.

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade