Case Study: Sticky.io recovers $8M in monthly declined transactions with dotData

How sticky.io Recovered $8M Per Month Using No-Code Automated Feature Engineering

sticky.io, a unified ecommerce platform processing billions in monthly transactions, was facing a significant challenge with over $23M in declined credit card transactions each month. Their data science team, led by Justin Shoolery, was overwhelmed by the immense and complex data from their ecosystem, spending the vast majority of their time on data wrangling instead of building models. They sought a solution that went beyond standard AutoML to help automate the most difficult part of their process.

The vendor dotData provided its dotData Cloud platform, which uses AI-powered feature engineering to automate the most time-consuming data preparation and feature discovery work. This allowed sticky.io's team to build a predictive AI model without writing a single line of code in a matter of days. The solution was highly effective, projecting a recovery of 35% of declined transactions, which equates to $8M monthly. As a result, sticky.io is expanding its data science team to leverage dotData for numerous other use cases.

Sticky.io

Justin Shoolery

Director of Data Science & Analytics