Customer :
Auto services retailer with stores nationwide.
Challenges :
A national retailer faced significant hurdles in accurately predicting store guest counts and the corresponding impact on Gross Sales, COGS, Net Sales, and Labor Hours. Their existing manual approach—primarily using spreadsheets—was time-consuming, prone to human error, and unable to incorporate complex data elements like inclement weather. Furthermore, they required an advanced forecasting model that could handle large volumes of data, including historical guest counts at 30-minute intervals and external data, to better understand the influence of external factors on store performance. No off-the-shelf solution met these highly specific requirements without extensive customization.
Solution :
To address these challenges, the retailer implemented Workday’s Demand Planning solution within Workday Adaptive Planning powered by AI/ML algorithms. The solution combines historical data from the customer’s Point of Sale system with historical and predicted independent variables, in this instance, leveraging predicted temperature data as a regressor to forecast future guest counts. By structuring data into cube sheets—including 30-minute increments—and creating a separate modeled sheet for weather inputs, they established a robust predictive framework.
Product mix assumptions (Product Factor) were then applied to the guest count ML forecasts, allowing the retailer to calculate volume. Which in turn drives Gross Sales, COGS, Net Sales, and Labor Hours. Additionally, Merge Cube sheets were leveraged to seamlessly consolidate different dimensions, eliminating the need for complex triggers and ensuring a streamlined data flow across various forecasting components.
Results :
This inaugural use of Workday’s Demand Planning, incorporating emerging machine learning capabilities, delivered transformative outcomes:
By integrating historical guest counts, regressor data, and product mix factors, the retailer gained a far more precise and nuanced forecast.
The previously manual, spreadsheet-based process was replaced by automated workflows, freeing teams to focus on higher-value activities such as strategy and analysis.
The organization’s unique data elements were successfully incorporated, offering a fully customized solution without the extensive development required by alternative products.
The retailer can now predict the financial impact of external factors like severe weather, enabling proactive resource allocation and store operation adjustments.
Conclusion :
By adopting Workday’s Demand Planning solution, this retailer achieved a significant leap forward in forecasting sophistication and operational efficiency. The integration of multiple data streams—ranging from historical guest counts to predictive temperature models—empowered them to plan for seasonal shifts and external disruptions more accurately than ever before. The success of this adaptive implementation underscores the power of machine learning-driven forecasting to tackle complex, mission-critical business challenges, paving the way for continued innovation and growth.
