Case Studies AI Demand Planning Saves Restaurant Chain $31M a Year

AI Demand Planning Saves Restaurant Chain $31M a Year

AI Demand Planning

A large fast casual restaurant chain with 660 stores facing over $50 million in annual labor expenses due to inaccurate store manager forecasts. Our demand planning system accurately predicts hourly demand for each store, resulting in improved estimates of consumer demand and an annual saving of $31 million for the company.

Critical Issue

This company was wasting over $50M in on annual labor expenses due to poor store manager forecasts regarding customer demand, which led to erratic work schedules for staff, resulting in either too much or too little coverage. This had the further effect of causing long customer wait times and numerous complaints.

Operation Icon

Customer Profile

Large fast casual restaurant chain

660 stores in 45 states

Crisis Icon

Key Problems

Over $50M of wasted annual labor expense on staffing empty restaurants

Long customer wait times and complaints

our solution

We developed a demand planning system that accurately predicts hourly demand for each restaurant and product line. This system utilizes weather data and local event data to make precise predictions. It operates as a fully automated process and is currently running in production. Additionally, we have created a user-friendly dashboard that allows users to monitor the accuracy of the predictions and provides intuitive explanations for the forecasts. Our implementation process was completed in just 12 weeks by a team of 3 individuals. The technology stack used for this system includes Azure, Databricks, and Python.

The results

Growth Icon

The company reported an annual saving of $31 million across 660 stores. Our innovative system has proven to be 30% more accurate than store managers’ forecasts, resulting in savings of $3,900 per month per store due to improved estimates of consumer demand.

hand holding a gear with workflow signifying support and management

Annual savings of $3.4 million in man-hours by enabling store managers to save 10 hours bi-weekly by not having to make forecasts. This has the added benefit of accelerating the effectiveness of new to industry store managers.

Award Icon

Our solution led to the accumulation of institutional knowledge as the drivers of the business are learned, encoded, and shared. Additionally, store managers have reported experiencing less stress from top-down pressure to predict the future.

Explore Our AI Solutions

Interested in exploring how our AI solutions are creating net-new revenue streams?

Concurrency Center of Excellence

Learn more about Concurrency Centers of Excellence online