Using machine learning and predictive analytics allows leading brewing company to improve logistics, operations, and customer satisfaction.

Founded more than 165 years ago, this worldwide brewing company has a portfolio of 100+ brands and carries a market share of more than 44% of the U.S. beer industry. With operations across the world—serving consumers of all backgrounds and interests and bringing them together over a shared love of beer—this company has a vested interest in making sure its customers are satisfied and that it is meeting consumer needs however they evolve. 

That begins at the front doors of the brewing company’s customers—retailers, bars, wholesalers, and distributors—as they are the gateway to delivering the end product to consumers. Understanding its customer’s needs and how to improve logistics, operations, and service is key to remaining successful and to building lasting relationships—especially in an ever-changing business landscape.  

Knowing The Pain Points Helps to Be ‘Hoptimistic’ About the Cure

Since the brewing company does business across the world, its customers and their needs vary. Pinpointing where there is dissatisfaction, and the reasons why, is not only critical to business sustainability, but it’s also necessary to preemptively address at-risk situations, support business growth, and provide best-in-class service to customers.

As part of the company’s commitment to outstanding delivery operations, the logistics division launched a survey to understand how its customers felt delivery drivers were doing. At customer accounts with unsatisfactory survey results, the worldwide brewing company would conduct visits to try and improve the relationship and avoid customer churn. The goal was to understand what the problem was, how it can be addressed, and hopefully—how it can ultimately be avoided.

The Biggest Challenge: ‘Barley’ Enough Data to Get Ahead of the Problem

The process to gather customer sentiment was very manual, which meant efforts to address poor ratings at many customer accounts were initiated too late. On top of this reactive process, the company was only getting a 10% survey response rate—and that did not provide enough context for the company to understand indicators that led to poor scores.

In addition to automating survey analysis, the worldwide brewing company also wanted to predict which customers were likely to churn instead of waiting for unhappy customers to share their dissatisfaction. If they could understand the drivers behind a poor rating, it could take proactive steps to improve delivery operations and customer satisfaction numbers.

The Solution: Company’s First Draft Survey Leads to Machine Learning Gold

The data that the company had previously collected—although not enough to see the full picture—didn’t go to waste. It was used to build a machine learning model that predicted how likely a customer is to give a bad rating.

The brewing company first developed a dataset containing nearly 50 features—individual, independent variables that act as input in your system which are used to make predictions—and more than 15,000 rows of data. The company then used Python to build a machine learning model from that dataset which identifies features that contribute to the probability of a bad driver rating.

The results of the machine learning model were then integrated into the company’s analytics tool so that the logistics team could build dashboards showing a ranking of customers at risk of giving a bad rating. The dashboards also provide a scorecard explaining which features contribute to the account’s risk score. The dashboard allows users to view influential features specific to an individual customer or across all accounts.


By looking at this data, the brewing company was able to see a few features that stood out as being the most predictive of a customer giving a bad score, including Off Day Deliveries, Missed Time Window, and Warehouse Breakage. Using this information, the brewing company knew what metrics to focus on improving and what accounts were most at risk.

The Results: A Dash of Data Science in The Brew Leads to Satisfaction Across the Board

Since turning to data science to help improve logistics, operations, and customer service, the worldwide brewing company has been able to take a more proactive approach to customer satisfaction and lowering customer churn.

By introducing machine learning and predictive analytics into its dashboards, the company has a more reliable list of accounts that may be at risk of churning, allowing team members to make visits to at-risk customers before they become real problems. The machine learning model also provides more insight into the “why” behind bad ratings and customer churn—enabling the company to think about solutions in a more thoughtful way, driven by data. Understanding specific KPIs that lead to unsatisfied customers allows the brewing company to offer personalized solutions and service on an account level as well as make overall process improvements that will positively impact the business going forward.


With more than 165 years in business, there is too much on the line not to get this right. This worldwide brewing company is now equipped to meet customer needs however and wherever they arise and provide the service they need, even before they know they need it.

Sharon Rehana Sharon Rehana is the content manager at Analytics8 with experience in creating content across multiple industries. She found a home in data and analytics because that’s where storytelling always begins.
Subscribe to

The Insider

Sign up to receive our monthly newsletter, and get the latest insights, tips, and advice.

Thank You!