As part of their commitment to outstanding delivery operations, a Logistics division at a Worldwide Brewing Company launched a survey to understand how their customers felt the company’s drivers were doing. At accounts with unsatisfactory survey results, the company would conduct customer visits to try and improve the relationship and avoid customer churn.
But the process to gather customer sentiment was very manual which meant efforts to address poor ratings at many accounts were initiated too late. On top of this reactive process, a 10% survey response rate did not provide enough context for the company to understand indicators that led to poor scores.
Instead of waiting for unhappy customers to share their dissatisfaction, the company wanted to predict which customers were likely to churn and understand the drivers behind a poor rating so they could take proactive steps to improve delivery operations and customer satisfaction numbers.
Using prior survey data, Analytics8 built a machine learning model that predicted how likely a customer is to give a bad rating.
We started by developing a dataset containing nearly 50 features and over 15,000 rows of data. We then used Python to build a machine learning model which identifies features that contribute to the probability of a bad driver rating.
We integrated the results of the model into the company’s analytics tool and built dashboards showing a ranking of customers at risk of giving a bad rating and 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 as a whole.
By looking at this data, the company was able to see a few features that stood out as being the most predictive of a customer giving a bad score which included Off Day Deliveries, Missed Time Window, and Warehouse breakage. Using this information, the company knew what metrics to focus on improving and what accounts were most at risk.
Machine learning models and advanced analytics techniques allow the company to take a more proactive approach to customer satisfaction and lowering customer churn.
By introducing predictive analytics into their dashboards, the company has a more reliable list of accounts that may be at risk of churning. This means they can make visits to at-risk customers before they become real problems.
The model also provides more insight into the “why” behind bad ratings and customer churn. Understanding specific KPIs that lead to unsatisfied customers allows the company to offer personalized solutions on an account level as well as make overall process improvements that will positively impact the business.
With stronger data and analytics, the company is equipped to improve logistics efficiency, delivery operations, and customer service.
To thrive with your data, your people, processes, and technology must all be data-focused. This may sound daunting, but we can help you get there. Sign up to meet with one of our analytics experts who will review your data struggles and help map out steps to achieve data-driven decision making.
In one hour, get practical advice that you can use to initiate or continue your move of data and analytics workloads to the cloud.
During your free one-hour cloud strategy session, we will: