Before
The grocery chain has its own bakeries where baking is made in-store from frozen semi-finished goods. Baker/Cashier decides which goods and how many need to be baked during the day.
It caused the following business challenges:
- The shelves display inconsistent product assortment, particularly during peak sales periods. Sales are declining as a result.
- Inability to promptly compile and analyze the actual sales plan for specific baking goods. As a result, it's difficult to make timely supply plans, leading to confusion in inventory management.
- Not optimal planning of baking cycles. Understocked shelves or increased write-offs as a result.
- Inability to effectively conduct promotions and analyze the result. For example, the pretzels appeared on the shelf only on the third day of Oktoberfest because the baker didn't notice the new boxes.
- Incorrect sales analytics. For example, low sales of croissants are not because of low demand, it was not just baked in the right quality. High sales of donuts do not mean their high demand. Sometimes customers could take them as a substitute for other more demanded goods.
A group of business analysts has evaluated the series of business hypotheses, the assortment matrix, and classified goods that are in demand at different times of the day. Their model was complex, taking into account the oven models, their capacity, the number of baking pans, and even the layout of specific units on the sheet. Product compatibility was also considered, you can't put garlic baguette on the same sheet with vanilla croissants.
To test these hypotheses, a customer needed a platform where they could quickly create a simple but scalable IT solution.
It could take up to 6 months to develop such a service on Dynamics NAV where most business processes in the company are implemented. The customer decided to pilot it on the Microsoft Power Platform.
Solution
2 Power Apps applications were developed in 3 weeks. One for bakers and one for analysts at the head office. Both applications share the same Dataverse database with a list of stores, goods, ovens, and baking modes.
Power Apps web application for analysts and marketers:
- Historical sales data view (transferring data from Dynamics NAV).
- AI model to forecast bakery sales.
- Setting up a baking schedule based on the store location.
- Goods compatibility constraints. For example, sweet and meat goods must be cooked separately.
- “Time of the day” baking schedule recommendations. For example, morning for croissants, noon - sandwiches, evening - bread and baguettes.
- Ability to quickly correct baking schedule in the exact store.
- A/B tests and hypotheses review.
- Goods stock forecasting.
Power Apps mobile application for the baker:
- Daily actions plan for a baker.
- Action details: good items and quantity.
- Completion status confirmation.
