COVID19 pandemic has changed the traditional paradigm of customer experience at the store and brought additional challenges to food and non-food retailers. Stores need to encourage social distancing, as well as a more "efficient" and fast purchase to their customers. Moreover, as the number of people allowed inside the store has become limited, a more efficient store can lead to increased sales. The ability to compute the "time in store" per customer according the shopping basket can assist in the identification of improvement opportunities both at commercial and operational level by:
In this analysis, we are going to tackle the problems mentioned above, leveraging the sales data from Sonae MC, the market leader in food retail in Portugal, with all the customers and products anonymised, as follows:
Let's start with the same spatial analysis presented on the previous post "1. Anomalies in Sales Between Stores", and use the API function SpatialMap() on the SKU "2003457". This function shows the spatial correlation with other products on the store (with a gradient showing how strong is the correlation), which in this case allows us to know that it is a non-essential product, and belongs to a specific customer segment that purchases 9 products. To understand how the location and the layout of the store influence shopping times, we created 2 different layouts and placed all products randomly. The results below show the spatial locations of the 9 products on each store layout.
Let’s now analyse the most likely path to get those products, and the shopping times, by leveraging our API function CustomerFloow(). The results below show the path of this customer segment, represented by ‘+’, considering the entrance is on the top right of the layout, represented by ‘0’, and exits also on the right. Knowing the paths of each customer segment, not only helps the retailer to understand the spatial correlation between products, but also the most likely direction the customer reach each product of the basket, which can be used to strategically choose the right location to add a new product or promotion. Studies show that when the shopping experience is highly personalised, customers indicated that they are 110% more likely to add additional items to their baskets and 40% more likely to spend more than they had planned. Moreover, when asked to rate a particular retailer, customers who experienced a high level of personalisation provided net promoter scores that were 20% higher than those of customers who experienced a low level of personalisation. Let's take a look at the results for the shopping times in both stores:
Comparing both stores, it is very clear that the left store is more efficient that the right store, in which the shopping time for the same basket on the right store (28min:09sec) is almost 2x bigger than the first store (18min:34sec). If the stores belong to different retailers, this customer segment will more likely go shop on the left store.
Feel free to reach us for any question or to try AssetFloow with your data. On the next post, we will explore the customer segments between e-commerce and physical stores, to get insights on how to leverage an omnichannel strategy to increase customer basket and create personalised experiences.
Source of the sales data and challenge description:
Whitepapers about AssetFloow intelligence applied to retail stores and warehouses.
#AI #Retail #DarkStores #SupplyChain #Warehouses #CustomerAnalytics