Intelligent Store Transformation – SAP Scan&Go Application with AI Based Recommendations
Part of the SAP AI Business Services product portfolio series. Credit to the co-authors Jens Mansfeld from the Retail Industry Business Unit and Steven Fu, Leonard Dinu and Utsav Garg from the AI Business Services team for drafting this blog post with me.
While the pandemic has accelerated e-commerce growth , it has not made the traditional brick-and-mortar stores obsolete. Physical retail stores are undergoing their own transformation to redefine the customer experience. Just as the pandemic has accelerated the adoption of automation and AI technologies, physical convenience stores are undergoing a transformation towards intelligent or unmanned stores. Unmanned stores offer the convenience of self-service and self-checkout, helping consumers save time in the shopping and checkout process. Retailers would be able to reduce cost in manpower overheads and can instead deploy automation and AI technologies to provide an omnichannel experience for consumers. The global unmanned convenience store market was valued at US$67.48M in 2019 and is expected to reach US$1,640.32M by 2027 with a Compound Annual Growth Rate (CAGR) of 51.9% .
How does the Scan&Go Application work?
In the following sections, we will describe the Scan&Go process flow as shown in figure 1 in a scenario with Anna (a retail customer) and Ralph (an unmanned store retailer).
Figure 1: Scan&Go Process Flow
Step 1: Shop and Scan
Anna is a regular customer at her nearby supermarket store which commonly has long queues at the checkout counter. Today she must prepare breakfast boxes for her family to bring to work and school the next day and decides to pick up a few items at a newly established unmanned store in her neighbourhood. After downloading the Scan&Go application, she enters the unmanned store, picks the items she wants, and simply scans the Barcode/QR on the price tag or the product itself by using a barcode scanner implemented into the mobile application. By doing this she is adding the product to her virtual shopping cart based in SAP Commerce Cloud.
Step 2: AI based Recommendation Service
In lieu of a store assistant, an AI based Recommendation Service embedded in the Scan&Go application provides a list of personalized recommendations to Anna for every item that she scans. The recommendations are derived based on the historical behaviour and item metadata, so that Anna and each shopper will receive highly personalized recommendations . With these recommendations Anna has the option to either purchase additional products to complement her shopping basket or choose a more favoured product among alternatives. From the retailer Ralph’s perspective the use of personalized recommendations to enhance the shoppers’ experience can increase both customer loyalty and boost their revenue through upsell and cross-sell their products In the example below in figure 2, Anna has picked the item super candy creamy peanut butter in the cart and received the recommendations for alternative brands of candies and an energy drink.
Figure 2: The Top 3 Recommendations from the Recommendation Engine
Step 3: Checkout Item
Finally, Anna can avoid the usual checkout queue and is able to checkout on-the-go by scanning the items using the third-party QR code and payment service. The point-of-sale (POS) transaction data collected here would constitute the operation O-data for further analysis by the retailer Ralph (see step 5 later).
Step 4: Qualtrics Survey
Normally Qualtrics  is used in Online Shops to get more information about the customers experience with the brand. Using Qualtrics in store to get information about the store experience is still a greenfield. Using a mobile application to shop enables this to the retailer. Today, when someone uses an uber, the passenger instantly receives a survey request after he leaves the car. This allows him to rate the driving experience immediately. This is a service that can be adopted to retail as well. See figure 3 below. When Anna pays for the basket an instant notification or email is send to her asking her about her shopping experience. This embedded Qualtrics survey will be send to Anna to ask for feedback and capture the customer, product, brand and employee experience insights all in one place. Ralph can also ask customers directly at the point of interaction why they chose to remove a product from their shopping cart. The survey results collected here would constitute the experience X-data using the Analytics Dashboard in step 5.
Figure 3: Qualtrics Survey to ask for User Feedback in the Scan&Go Mobile Application
Step 5: SAP Analytics Cloud
Furthermore, SAP Analytics Cloud enables Ralph to analyse all their data based on advanced text-based and statistics-based analysis along with predictive modelling. The information obtained is then send to the SAP Analytics Cloud where operational O-data like sales data of the retail stores is merged with the experience X-data of the Qualtrics survey. This allows Ralph to find correlations between the store experience and the store performance on every level of detail. With real time shopping records and online surveys collected, the SAP Analytics Cloud dashboard provides a summary of the retailer’s performance as shown in figure 4. The dashboard comprises two sections. The operational O-data on the left display’s sales figures, average cart value, average item value and more. On the right the experience X-data summarizes the customers’ review on their shopping experience.
Figure 4: SAP Analytics Cloud Displays both the Operational Sales Data and Experience Data for Effective Management Oversight
The Omnichannel Retail Experience
The pandemic has changed people’s lives forever, and the way people work, live and shop. This presents a golden opportunity for retailers to undergo retail transformation. The Scan&Go mobile application from SAP enables the rise of unmanned stores and provides a unique in-store experience for every shopper. Mobile Applications in store, in combination with customer loyalty programs allow retailers to serve the segment of one. Each customer can get individual services like personalized recommendations or promotions based on his/her shopping history and his/her preferences. In combination with SAP Analytics solutions this allows retailers to get a detailed overview about what is happening in their store, that is not POS (Point-of-Sale) data related.
Check out the demo video of Scan&Go mobile application here. The video also describes the detailed explanation of the AI Explanability and the overall Scan&Go architecture.
Retail Industry Business Unit, Solution Specialist Digital Innovation: Jens Mansfeld
AI Based Recommendations Product Manager: Kwok Peng Lai
AI Based Recommendations Product Owner: Steven Fu
AI Based Recommendations Machine Learning Developer: Dinu, Leonard and Utsav Garg
- Find more information about SAP AI Business Services
If you have more questions feel free to reach out to us or ask your questions over SAP Community using the hashtag #Artificial Intelligence.
Read all blog posts of the SAP AI Business Services introductory–, and product portfolio series