Beverage Case Study
The sales volume of beverages is highly sensitive to seasonal and weather variations, which can be challenging to predict and manage. By utilizing AI and statistical methods, the company creates and operates a sales demand management model tailored to its client companies. This model is able to analyze and predict changes in demand, which allows the company to optimize inventory levels and prevent out-of-stocks and excess inventory. By using advanced analytics, the company is able to improve its forecasting accuracy and minimize the impact of external factors, such as weather and seasonality, on its sales. As a result, the company is able to provide its clients with a reliable and efficient service, which in turn helps them to manage their inventory more effectively and improve their overall business performance.
|Business description||Manufacture and sale of soft drinks|
|Introduction module||Demand Forecast (AI & Statistics) | Sales Plan | Replenishment Plan | Supply plan | Production plan | Inventory management | Analysis & Report|
"Company A", a leading beverage manufacturer that specializes in make-to-stock production, is facing challenges in effectively forecasting demand for their hundreds of SKUs. The current method of using Excel results in low accuracy predictions, making it difficult to manage inventory properly. This leads to issues such as out-of-stock situations, excess inventory, and expired products. The short expiration date of the products further compounds these challenges. Additionally, the company frequently receives urgent orders from large supermarkets and convenience stores due to sudden promotions, which require significant time and effort to adjust production plans accordingly.
Incorporating weather information and sales promotion data into the demand forecasting process, along with the use of AI (machine learning) can help prevent shortages and improve forecasting accuracy for Company A. The AI engine takes various input data such as meteorological information, holiday information, economic indices, and discount rates, into account. T³SmartSCM's BF module utilizes machine learning as a key method for demand forecasting to make more accurate predictions, resulting in better inventory management and reduced waste due to overstocking or stockouts.
The company compares the performance of conventional forecasting methods with that of statistical models and AI models, and automatically selects the most accurate data as reference values for sales planning. This improves the prediction accuracy and helps the company to make more informed decisions. The system automatically presents the most accurate data from the three cases of planning with Excel, statistical model, and AI model as a Baseline Forecast (reference value) for sales planning. This ensures that the company is always using the most accurate forecast data for sales planning, making it easier to optimize inventory, reduce waste, and improve overall efficiency.
When formulating a sales plan based on the demand forecast, the company can take into consideration the sales promotion data of its business partners (retail stores) to make more informed decisions. This allows the company to adapt its plan in a timely manner to changing market conditions, such as special promotions or events, ensuring that the company can always meet customer demand. Additionally, by integrating the sales promotion data of the business partners, the company can better align its inventory and production with the specific needs of each retail store, resulting in more efficient use of resources and better customer satisfaction.
To effectively manage the thousands of products in its inventory, Company A categorizes them according to their sales characteristics, customer preferences, and product importance. By dividing the products into dozens of product groups, the company can implement tailored sales strategies and safety inventory policies for each group. This helps prevent shortages and excess inventory, as the company can adjust its inventory levels according to the unique demands of each product group.
Furthermore, the company optimizes its inventory by managing the amount of inventory according to the inventory grade. This ensures that products with a higher importance or higher customer demand are prioritized, while products with lower demand are managed accordingly. This approach enables the company to maintain optimal inventory levels, reduce waste, and improve overall efficiency.
Return on Investments
By implementing a replenishment plan and inventory management plan based on the rules for each product group and the demand plan for each product, Company A is now able to effectively prevent out-of-stocks and excess inventory. This approach allows the company to optimize inventory levels and maximize profits without losing potential sales opportunities.
Automating demand forecasting has enabled the company to more quickly and accurately manage inventory calculations by item and business partner, replacing the previous Excel method. This, in turn, allows the company to formulate a realistic production plan for each factory, taking into account all constraints such as setup changes, available inventory, and resource capacity.
Furthermore, the company can now handle manual changes caused by urgent orders from promotional events more efficiently, ensuring that inventory and production can be adjusted in a timely and effective manner.
The supply chain planning system can simulate various scenarios involving various raw material supplies and their prices, and come up with the best scenarios that would maximize profits while keeping the sufficient supplies to meet the demand.
Baseline Forecast (AI & Statistics) | Sales Plan | Supply Plan | Factory Plan | Analysis & Report