For [Redacted]’s convenience stores, it is imperative for the company to have accurate sales models for the inventory held at each location. Numerous factors affect individual stores, or groups of stores calculations of these forecasting models, including location, size of the store, day of the week, holidays, weather, and many more variables. [Redacted] also has the challenge of determining whether they are fulfilling the needs of the customer by carrying the items the [Redacted] customers want. For this project, [Redacted] would like to improve the sales forecasting model of candy bars. Currently, [Redacted] does not have a model for forecasting item-level sales in these categories. A member of the [Redacted] team will be building out a model simultaneously and at some point will provide an update on how their own model is going.
[Redacted] has provided 1.4 million rows of historical sales data beginning in September 2017 through August 2020 and expects the model to be built using this historical data set. The project goal is to build a forecasting model that can be easily integrated into [Redacted]’s current workflow and aims to improve their current system of managing candy bar forecasts by enabling better management of candy bar:
• Inventory
• Pricing
• Budgeting
• Revenue forecasting
In order for [Redacted] to generate accurate budgets and financial forecasts at their convenience store locations, sound forecasting models must be built for the different items they carry in the stores. In this project, the goal is to build an accurate forecasting model of candy bar sales.