Credentials: CINDE: symphonyretailai.com account Assortment: symphonyretailai.com account Store Intelligence: demouser / srai2024 Demand Forecasting: gb/gb CINDE Merchandising Sales Driver tree – highlight and summary CINDE even summarized for us that [Show the narrative] “sales are down due to a decrease in trip size driven by customers shopping less broadly across categories”. Straight away, we want to understand which categories are being dropped from customers’ basket. To do so, we can simply ask the Copilot a question: • Which categories have the worst decreased basket penetration? Soft drinks and corn chips are the two worst-performing categories. This requires a deeper dive. Take corn chips, for example. • Let’s review the corn chips category over 52 weeks • Looking at my competition, what prices should I change? • • How can I improve promotions in the corn chips category? Let’s ask the Copilot: • Which products should be added and delisted in the chips category? https://demo-catman-xcelerate.insights2execution.com/platform We will now optimize the assortment for our Xcelerate store and the shoppers who shop there… Cluster • Through Intelligent Clustering we will identify that there are two significant preferences when buying chips. Shoppers generally have more traditional tastes in flavours or tend to be more adventurous. • Those shopping in the cities are more adventurous, whilst those in the more suburban areas stick to the traditional flavours. • Of course, this could correlate with other factors like families buying for children in the suburbs with plainer preferences. Assort • Our NRF shoppers prefer spicy chips and world flavours aligning with their more adventurous preferences. We will align the assortment to these preferences. • Our NRF store is small with only a few products, so the system correctly shows us that the sales of most products are highly incremental. Dynamic incrementality calculations at the store level are essential for accurate assortment analysis. • We will DELIST two products that are poorly performing both in terms of sales, volume and loyalty. • We will ADD two products that are likely to perform strongly and drive increased sales and larger baskets together with long-term customer loyalty. The key point to communicate is that our loyalty data is at a store level – no one else does that! • We will analyze the sales, volume and margin impact of our decisions. • We will create a new planogram with our space-aware assortment and push the new assortment and planogram to the Electronic Shelf Labels. 4. Instore Profiles Store Intelligence SI >> ESL >> Mobility & Refreshment Store Intelligence and Electronic Shelf Edge Labels Since it’s difficult to do this without a shelf setup we will do the training on Saturday afternoon 4PM and Sunday morning 8AM at the show. Please download the Store Intelligence app to your phone. Instance: CPGD Username: Xcelerate / Password: ret3ch2024 Store: Xcelerate 2024 When you log in to the Store Intelligence app you will find the following stores: • NRF Fixed Cameras • NRF Mobile • Hanshow Fixed Cameras • Hanshow Mobile • Data Logic Mobile VERY IMPORTANT: DO NOT TAKE CAPTURES ON THE MOBILE APP FOR CAMERA STORES. ONLY USE “MOBILE” STORES FOR MOBILE CAPTURE. Ensure to select the correct store for the stand you are on. You may also wish to show the results of captures on the big screen, Use the following URL and the same login details as above: https://cpgd.rebotics.net/ Key Story Points • Promotions, Assortment changes, Price and MDM data will all be pushed automatically to ESLs. • Store Intelligence will capture images through a fixed camera and push the actions to the shelf labels in real-time. This directs the store staff exactly what to do to create a perfect and efficient store environment. Actions include: ◦ MOVE a product ◦ REMOVE an invader ◦ Replenish an Out of Stock product from backroom stock • FreshAI – we will monitor stock levels of fresh items through fixed cameras and automatically create tasks to replenish fresh items from the back room. Copilot demonstration (Generative AI for Demand Forecasting) The idea of the copilot to help the demand planner is again to help understand complex scenarios and questions. We will walk through the what, why, and how: “Step 1: Show me what is happening, Step 2: Help me understand, step 3: How do we move forward” USECASE: 1 (FORECAST EXPLAINABILITY ON IMPACT FROM PROMOTIONS Tying Forecast to Promo Planning: promo data from promo planning feeds to DFAI to forecast impact from promotion; THIS SCENARIO WOULD BE AFTER A PROMO HAS ALREADY BEGUN IN A STORE and there has been a high impact to DC shipments to the store, AND THE FORECAST IS LEARNING OVERTIME WHAT IS HAPPENING, WHY, AND HOW TO CORRECT) PROMPTS: Step 1: Show me what is happening (the WHAT) Show me items that have significant forecast inaccuracies due to unusually high actual shipments Vs forecasted in DC 90015. Step 2: Help me understand why this is happening (the WHY) Let's focus on the 3rd line. Help me to understand the reasons as to why this item resulted in very high actuals shipment qty than forecasted. Step 3: Help me move forward (the HOW) I would like to adjust the high actuals for this item in order to use it in the upcoming forecast run. Could you please help with that ? USECASE: 2 (NEW ITEMS INTRODUCED TO THE ASSORTMENT – FORECAST ANALYSIS AND AI SELECTION FOR LINKING A LIKE ITEM – TIEING THE FORECAST TO ASSORTMENT) Assortment Optimization: new or removed items from assortment are triggered to SCM to update the MDM (indicating updates to authorized items) and to DFAI to indicate what needs to be forecasted (new items) and/or end the forecast and replenishment for removed items PROMPTS: Step 1: Show me what is happening (the WHAT) Show me the list of newly introduced items in DC 90015 Step 2: Help me understand Let's focus on 1st record. Please suggest up to 5 similar items within the same category. Please show me the actuals of these items in the application. (or you can click the suggested button “View actuals” blue button in video…) Step 3: Help me move forward (the HOW – take action) Can you recommend an item that is best fit to link from this suggested list?.