In the context of this funding project with the Fashion Tech Start-Up Dresslife, a system is designed, that aims to solve the problem of low conversions and high return rates in fashion e-commerce. By applying deeplearning methods, the information asymmetry of customer, product assortment and company will be reduced. In addition, the technological solution improves the environmental impact of global fashion e-commerce because the avoidable returns cause CO2 emissions.
One of the biggest cost drivers in fashion eCommerce is return shipments - and processing costs due to a high rate of returns, because buyers either do not like the clothes they have ordered or they do not match their personal taste. The ordered goods are sent back and return rates of up to 80% occur. When the returned goods arrive at the company, the package is unpacked by the employees, the condition of the goods is checked, the goods are cleaned / reconditioned if necessary and repacked. In addition to the resulting process costs and the tying up of resources, there is often a loss of value due to the use or testing of the customers. The effects include reduced profit margins of online retailers / fashion companies, unnecessary transportation, increased CO2 emissions, wasted packaging material, inefficiencies in business processes, dissatisfied customers and a poor external image of the company.
The development goal includes the unfolding of A.I. technology on the customer and company side in order to ensure a profound development of the keep rate through deep learning. By maximizing the keep rate, we can reduce the return rate by at least 50% and significantly lower process costs. In addition, the conversion rate in the online shop is maximized and the flexibility and speed of innovation of the fashion retailers is improved.