Customer Segmentation utilizing Data Science and Machine Learning within Incorta

Summary

Having deeper insight into customer types is a competitive advantage across all markets. One example is targeted marketing, where effecting campaign strategies can be tailored to one specific audience at a time. Customer segmentation enables businesses to customize their decision-making processes to different customer subgroups. There is also no need to know exactly how many segments exist beforehand, as machine learning will optimize how segments are defined based on the data being fed into the algorithm.

• Choose what data is relevant when segmenting customers
• Trust in the output of a high-performance machine learning algorithm
• Understand segment profiles and tailor business strategies to each segment

Components

Data Processing

• Aggregated customer-level views
• Data validation checks
• Scheduled data loading

Feature Engineering

• Transactional metrics
• Normalization and vectorization
• Pipeline tuning

Machine Learning

• Unsupervised learning framework
• K-means clustering
• Cross validation feedback loop

Dashboard Insights

• Segment profiles
• Metric correlation analysis
• Top products by segment

‍