Why can't I compare the Return Rate due to size with and without Usizy, in the dashboard?

Directly comparing the return rate of orders placed using Usizy against those placed without it does not yield a valid analysis[cite: 1]. Visitors who interact with our service before purchasing exhibit completely different buying behaviors and come from a fundamentally different context than those who do not[cite: 1].

Understanding and analyzing return patterns involves several complex variables that distort a direct numerical comparison[cite: 1]:

Key Analytical Challenges

  • Variations in Visitor Profile: Direct visitors often include loyal or returning customers who are already highly familiar with your brand’s sizing and do not need to verify their fit[cite: 1]. Consequently, this specific cohort naturally exhibits a lower return rate[cite: 1].
  • User Discretion: A portion of Usizy users choose to manually override or ignore our final size recommendation during checkout[cite: 1]. When these users subsequently experience a poor fit and return the item, it reflects user preference rather than an inaccurate system recommendation[cite: 1].
  • Multiple Profile Adjustments: Some visitors modify their initial body measurements during a single session, receiving more than one size recommendation for the same item[cite: 1]. If a purchase follows, it is impossible to definitively track which of the multiple recommendations the customer ultimately relied upon[cite: 1].
  • Inherent Return Data Incompleteness: It is practically impossible to guarantee when a complete, absolute list of returned items has been successfully captured for any given period[cite: 1]. This unpredictability stems from several operational factors[cite: 1]:
    • Extended Return Windows: Depending on your store’s return policy, customers may have 30, 60, 90 days, or longer to return an item, creating a significant data lag[cite: 1].
    • Data Transmission Gaps: We may not receive all returns due to technical issues or manual files not being regularly sent.[cite: 1].
    • Misclassified Return Reasons: Size-related returns are frequently mislabeled by the end customer or warehouse staff, remaining permanently flagged under an incorrect return category[cite: 1].

To prevent flawed data interpretation, we deliberately omit this specific comparison from the dashboard[cite: 1]. Instead, your integrated returns data is utilized on the backend to fuel our Machine Learning models, driving the continuous optimization and improvement of our core size recommendation engine[cite: 1].