Case Study: Flo improves irregular cycle prediction accuracy by 54.2% with InData Labs

Flo improves cycle prediction accuracy by 54.2% with InData Labs

Flo, a smart period tracking application, faced the challenge of accurately predicting menstrual cycles for its large user base, particularly for the approximately 30% of women with irregular cycles. Their existing method, based on an average formula, resulted in a 5.6-day error margin and a loss of trust from many users. To address this, Flo partnered with vendor InData Labs for neural network implementation and predictive modeling.

InData Labs developed a sophisticated neural network that analyzes over 400 individual user inputs to create personalized predictions. This solution improved prediction accuracy for irregular cycles by up to 54.2%, reducing the error margin from 5.6 days to just 2.6 days. Following this implementation by InData Labs, Flo became the most downloaded app worldwide in its category, significantly enhancing user trust and satisfaction.

Flo