I identified a niche product feature and analyzed the usage of all key functionalities by cohort to discover it was only being used by 12% of its potential audience. I then deprioritized all engineering work focused on enhancing that feature, which saved $45k in unnecessary development costs over two months.