Authors:
Jessica Snead, Susan Brosnan, Daniel Goins, Patricia Hogan
Abstract:
A method is proposed to help an online clothing retailer to better manage their inventory and provide customers incentives as part of that method.
Background:
When a customer uses an online wardrobe service, they purchase multiple items with the intention of keeping only a few. This impacts the retailer's inventory and the retailer's plans for returns processing. How can the customer be incented to return un-wanted items sooner, or to not request so many items when they know they will only keep a few and return the rest?
Description:
Wardrobe Inventory Management Use Case:
A customer uses a wardrobe service to purchase several items, with the intention of returning many, but they want to try on all the items at home and make a selection. If the customer purchases 10 shirts of different styles, sizes, and colors and may keep 2-3 of them, then the wardrobe service has a tricky inventory problem - with the unknowns of what items a customer is likely to keep vs return and when. The wardrobe service cannot offer the items that will be returned to other shoppers until they items have been returned.
The following is a list of things the shopper could be encouraged to do, to help the retailer have a better idea of which items will be returned.
- Customer can indicate how many items from each grouping they intend to keep - to help the service with inventory and returns planning
- Customer can get an incentive to return items more quickly for a bonus or discount
- Customer can earn more items allowed to "borrow" at a time, or more time until return is due by submitting photos and reviews of the shopper wearing the items.
- Customer can earn reliability points by accurately reporting their intentions and quickly sending returns, which can be exchanged for discounts or benefits.
On the retailer’s side the retailer could use machine learning and analytics to figure out which items the shopper is most likely to keep and most likely to return. They could learn from the customer's buying history or from similar customer's buying history and behavior.
-- Retailer / e-tailer knows that Person X usually buys size 8
-- Person X orders size 6, 8 and 10
-- Retailer could send Size 8 immediately, and delay sending the other sizes
-- Reserve + Queue up Sizes 6 and 10, giving time for Person X to try Size 8
-- If the shopper likes size 8, the shopper could tell the retailer to cancel the other two sizes before they were sent, saving the e-tailer money/hassle, perhaps the shopper receives points or coupons for responding promptly
-- if the shopper orders the same item but in different colors, the retailer could use analytics to predict which color the shopper is most likely to buy based on the colors the customer tends to buy, or other shoppers tend to buy. The retailer could send the color they think the shopper is most likely to buy first and delay sending the other colors. A similar technique could be used with styles. If the retailer knows the shopper tends to buy a certain style or look, or length of garment, the retailer could send that type of item first.
The retailer could also use analytics to pick the size, color and style the shopper is most likely to like, and then if the shopper requested other sizes, the retailer could send the other sizes in other colors, that are maybe less popular so that the retailer is not tying up items in the most sought-after colors. The sought after colors would be available to other shoppers. The first shopper can try different sizes and see how they fit, but they are likely to return certain sizes.
The main thing that differentiates our idea from existing art is the part where the retailer uses analytics to send a subset of the items the shopper selected. The retailer sends the items, the retailer things the shopper is most likely to buy. The retailer delays sending items in the bracketed set that the retailer thinks the shoppers is most likely to return, especially if those items are items that are in high demand by other shoppers.
TGCS Reference 2965
