Case Study: Major Retailing Company boosts online sales by revealing customer moods with Teradata

Retailer Analyzes Unused Data to Reveal Customer Moods, Boost Online Sales

A major online retailer had large volumes of unused, time-stamped transactional log data and wanted to better understand the “hidden states” — customers’ moods and contextual behaviors that drive purchase, impulse buys, or abandonment — so it could improve conversion and personalize real-time engagement.

The retailer engaged Think Big Analytics (a Teradata company) and used the Teradata RACE proof-of-value to move Hadoop logs into Teradata Aster and apply Hidden Markov Models to sequence customer actions. The analysis revealed distinct behavior patterns and purchase probabilities, enabled targeted contextual offers and re‑engagement strategies, and produced visualizations (the “Star”) to map customer journeys—delivering improved online sales and insights that are applicable across industries.