How to reduce eCommerce returns by 60% (YC S22)

Hey HN, co-founder of REVER (YC S22) here! REVER is helping brands manage eCommerce returns and make them more profitable. 6 months ago, we got ambitious and thought: what if we can actually reduce returns? Not just make them easier or more profitable but actually reduce them by avoiding them before they happen. So as usual, we just built it: REVER size suggestions. A way to reduce returns by suggesting size insights at checkout using AI. We know what people return, why do they return it, how does this change by country and what do they prefer instead. So we built a model that identifies the most returned products & their reasons and use this to help shoppers avoid sizing mistakes while buying. We fine tuned it with market insights and found out that this was specially relevant split by country: most of the times the same product was an outlier for "Too small" in Germany and for "Too big" in Portugal. Bodies are different across geographies, so sizes should too. So now, when a shopper is buying in a brand using this feature, we identify the market and based on returns data, we show the corresponding size suggestion. The result? Returns reduced by +60% for brands like HOFF on the affected products and the feature is now out of beta for all +500 customers.