Umbrella Co runs a B2C e-commerce operation with 2.3 million active customers and a catalog of over 40,000 SKUs. Their email personalization program was generating recommendations from a weekly batch job that ranked products by historical purchase affinity.
The Problem with Batch Personalization
A customer who browsed winter coats on Monday and received a personalized email on Thursday would see recommendations based on last week's behavior, not their current intent. For seasonal products and fast-moving inventory, weekly recommendations were frequently irrelevant or out-of-stock.
Building Real-Time Behavioral Segments
The team used MarTech Pulse to define behavioral segments based on browse and cart events from the last 48 hours. These segments fed directly into their email service provider as dynamic audience conditions, so every email send pulled fresh membership at send time.
Outcomes
Within 60 days of going live, email CTR increased by 28% and average order value from email-attributed purchases increased by 15%. The team attributes this to the behavioral relevance improvement — customers receiving recommendations for products they had actively browsed in the previous 48 hours, rather than historical affinities.