Background: adidas is the second largest sportswear manufacturer in the world.
Adidas is the second largest sportswear manufacturer in the world, and they opened the first store in China in 1980. They have more than 8,500 stores in China including distributors right now.
Do not know the report that distributors turned in is real or fake
Adidas has many stores in China, so it is difficult for them to manage each store. Therefore, for managing stores convenient and easy, adidas divided all stores from A to E. A means best-selling performance stores, so they have the priority to import new trending products, and they also have more opportunities to get promoting by marketing department. What a wonderful group. Because group A has too many benefit, all distributors would try their best to make a great selling report no matter that one was real or fake. Besides, some distributors will beg their friends who work in adidas to help them. Because of these different factors, adidas could not identify the real group A stores, and they might estimate the wrong number of order to the stores. When a fake store of group A received the orders that exceed their sales ability, this store would start to do some discount for attracting people to buy products. If all adidas products were usually on sale in this store, clients would not willing to buy them in original price gradually. One day, this brand, adidas, will become a second-level sportswear manufacturer. Thus, adidas needed a method for helping them to identify the authenticity of reports to avoid this disaster.
Main Solution Methods we used in this case:
Kiwi, a location intelligence platform
getchee decided to construct a systematical standard procedure for helping adidas to redefine the group A to E, and estimate the revenue of all stores to identify the reports that distributors presented are true or not.
We used demographic segmentation to solve this problem. First, segment all stores by city tier and region. Second, distribute total population, working population, household number, and age. Third, model the distribution of wealth. Fourth, target key competitors. Fifth, group the location and the capacity of each store. Final, compile customer historical purchase behavior and find out the purchase trend. After analyzing all different factors of each stores, we combined and applied all results to our platform, Kiwi, and presented a visual score card to adidas. They can easily realize the current situation of each store, and export the proper quantity of stocks. Now, adidas can control the export volume by themselves, and they also avoid the disaster, being a secondary brand.