Project Description
SKU-Level Forecasting Accuracy Drives Strategies for Precision and Growth
This leading spirits company achieved a 15-point improvement in forecasting accuracy through the implementation of AI-driven SKU and regional-level demand planning capabilities.
Client Challenge
The client is a leading premium wine and spirits company, marketing and distributing the most comprehensive portfolio in the industry with 240 brands in over 160 countries. The company operates through a network of regional distributors, with widely varying retail store types and distribution channels across the United States. Product planning must factor in different state laws for controlled substances, including 17 ‘control states,’ where alcoholic beverages are purchased through ABC boards in each state. As such, transaction data comes in varied formats and timeframes.
The COVID-19 pandemic drove massive fluctuations in product depletion across retail stores (off-premise), as well as consumption in restaurants and bars (on-premise). This severely reduced forecasting accuracy and resulted in significant inventory shortages and overages nationwide. Further, the company’s depletion forecast was only available at the distributor level and did not provide visibility into long-term production planning.
The company wanted better forecasting capabilities to improve product planning and distribution for each region in the U.S. on an item-level basis.
Parker Avery Solution
The Parker Avery Group’s analytics team employed the firm’s advanced demand analytics solution to drive an accurate sales mix and a stable, low-biased long-term forecast for the company.
Key factors to Parker Avery’s solution included:
- Market/SKU-level forecasting
- Time-series analytics
- AI/ML predictive analytics
- External data sources (COVID-19 database, macroeconomic variables, etc.)
- Basic pricing
- Long-range forecasting (18 months)
Further, the solution focused on management by exception to reduce manual intervention by planners and enable the inclusion of insights to which analytics does not have visibility (e.g., supply chain constraints).
Results
The client’s forecasting accuracy has improved by 15 points across the company’s brands, and they can now forecast down to the SKU (item) level across all geographies. The new forecast is seamlessly integrated into existing systems to optimize demand planning.
Further, the solution provides a single, holistic demand signal that the company will leverage for strategic planning (18-month forecast), sales and operations planning (S&OP), and trade promotion management (TPM). Parker Avery continues to work with the client to improve forecast accuracy through modeling improvements such as data harvesting, price promotion modeling, and the incorporation of additional demand drivers (e.g., ad spend).
You may also like
