Authors:
Julia Patten
Abstract:
Retailers don't often have data points to use to understand customer decision patterns with respect to the customer's location in the store when items are added to or removed from shopping carts.
This disclosure proposes adding location data for the carts in conjunction with item addition and removal data to provide additional data points to use for alerts and store layout and marketing insights. The goal of these insights are to increase basket size and reduce shrink.
Background:
Alerts can be derived using cart location data in addition to item addition or removal which can be used to handle real-time re-shelving needs.
Alert examples:
* A frozen item was removed from the cart in another aisle. After receiving an alert, an employee can ensure that the product is re-shelved properly
* A daily list of items that need to be moved back to their original position can be created; customers may not purchase items from a location different from their home shelf as there is no way to compare products. (This theory can be tested with the data points provided)
In addition, data can be used for product placement and marketing.
Marketing examples:
* How effective is product placement in other locations such as an end-cap?
* Are item swaps done here, adding an end-cap featured brand and removing another brand of the same item which was picked up from its usual location?
* How do most customers normally traverse the store aisles on a shopping trip?
Description:
This is a light-weight event driven data point capture for immediate and future analysis. The data collected can be just one feed used in a data lake to use to analyze and possibly understand shopper’s patterns when removing items from their cart.
Unlike a system of shelf-mounted cameras which can be used for hand-detection and view data, the cart can be located by sensors on it or even possibly using the customer location data through a retailer’s app when a customer has checked-in.
The data is point in time, ‘where was the customer when they took ‘x’ out of their basket?’. It does not require constant monitoring of the customer’s path or of their facial expressions, but this feed can give us data points such as:
- ‘where was this item removed’
- ‘was another item added when it was removed’ (which might be a brand exchange based on the SKU removed and the SKU added)
- ‘was this the last shelf/endcap on the customer’s way to checkout’
In fact, data addition/removal event data analyzed in order relying on subsequence can even lead to decision patterns by itself with no location data required.
In essence, combining two different technologies increases the data available for retail analytics.
Knowing both where a cart is in the store in addition to the customers item adds/removals from the cart at that location can provide valuable data on customer shopping behavior. This lightweight feed of location data paired with item additions and removals without requiring massive camera feeds, image analysis, or mathematical analysis can provide real-time data points for shopping behavior.
Claims:
Both location data paired with item addition/removal data can be captured in real time and acted on (ex: ice cream removed from the cart in the wrong aisle), in addition to being used for analysis of patterns.
Enabling Technology:
This idea is concerned with shopping carts having the ability to detect when items are placed into or removed from carts.
https://patents.justia.com/patent/11059506
Cart systems that track item insertion and removal are primarily interested in ensuring the transaction contains the correct items (leading toward touchless checkout systems).
TGCS Reference 2832
