This subscription box brand was manually assigning monthly shipments using hand-written rules, which introduced bias and limited personalization at scale.
A leading subscription E-commerce company replaced manual product assignments with Faraday’s algorithmic matching—lifting revenue and retention through smarter personalization.
To improve performance, they tested AI models from Faraday that provide custom predictive datapoints for a two-sided matching approach—scoring both customer and product attributes—to predict high-likelihood pairings and deliver the most relevant and value-driving recommendations.
A rigorous A/B test showed Faraday’s algorithm drove a ~5% lift in revenue per customer compared to the rule-based logic.
The algorithm also reduced churn by ~3% and generated an over 5x monthly return on investment.
Faraday’s recommendations outperformed manual rules for four consecutive months—and the brand continues to experiment with new strategies to drive further gains. Moreover, these recommendations were delivered automatically to their CRM as ready-to-use custom datapoints.
