About the project
Initial analysis reveals that fashion returns are not driven by fit and size alone - fabric feel, quality, and real-world expectations play an equally important role. It is also observed that shoppers place greater trust in feedback from other buyers compared to purely AI-led suggestions. Based on these insights, the solutions are designed using Myntra as the base platform, focusing on strengthening human-led signals and improving product understanding before purchase. The following solutions aim to equip shoppers with clearer, more reliable information to make confident decisions and reduce mismatch-driven returns.
Return Insights transforms collective shopper behavior into confidence cues
Research showed that shoppers trust peer experiences and real purchase outcomes more than algorithmic recommendations. To reflect this and avoid misleading signals, Return Insights are shown only after a minimum purchase threshold is reached (e.g., 150 purchases), ensuring the data is reliable and meaningful.
Products are then grouped into four confidence states from new items with insufficient data to low-return products translating return behavior into simple, actionable cues. Rather than exposing raw numbers, this approach helps shoppers know what to double-check before buying, shifting returns from a post-purchase issue to a pre-purchase confidence signal.
Detailed Reviews as a Lever to Reduce Returns
Analysis reveals a clear inverse relationship between review quality and return rates products with more detailed, structured reviews experience fewer returns due to better expectation setting. To address this, a detailed review sheet is introduced, guiding buyers to share feedback across key return-driving attributes such as fit, fabric, size accuracy, color, and quality.
A 5% product-value reward incentivises complete reviews across ratings, text, photos, and videos. A clear progress tracker ensures rewards are visible, transparent, and motivating.
