Explore how Nihilent used Resense to help one of India’s largest quick service restaurant chains forecast daily sales, orders and ingredient consumption for more than 1,300 restaurants, balancing stock outs against kitchen waste.
Service
Data & Analytics; Resense
Vertical
Energy
Region
Asia
Tech Stack
Resense Demand Sensing and Forecasting, 40+ Time Series and Machine Learning Algorithms, 12 External Predictors Including Weather and Local Events, What If and Sensitivity Analysis
The client is one of the largest quick service restaurant chains in India, with more than 1,300 restaurants. Its supply chain was already efficient, but restaurants and suppliers had no unified view of supply and demand. The result was waste in the kitchen, unconfirmed supply of ingredients and, at worst, a broken customer experience.
Restaurants ordered ingredients without a reliable view of what each kitchen would sell the next day, creating waste and shortages at the same time.
01
Over ordering meant perishable ingredients were thrown away.
02
Under ordering meant restaurants ran short when customers were waiting.
03
Restaurants and suppliers planned separately, without a unified picture.
04
Weather, local events and promotions changed demand restaurant by restaurant.
05
New stores, split stores, new ingredients and recipe changes all disrupted the historical pattern.
Nihilent deployed Resense as an auto indenting solution, forecasting at the level of each restaurant, each ingredient and each day.
Auto Indenting for Every Restaurant
Built an automated indenting solution for restaurants across India.
Granular Forecasts
Forecast daily sales, orders and consumption of more than 300 ingredients for each restaurant using more than 40 time series and machine learning algorithms.
External Demand Drivers
Integrated 12 external predictors, including weather and local events, to refine accuracy, and measured the impact of each driver.
Safety Stock by Volatility
Set safety stock according to demand and supply volatility to balance stock outs against waste.
What If and Promotion Analysis
Modeled the effect of service levels on inventory cost, stock outs and waste, and the sensitivity of demand to promotions.
Built for Change
Adjusted forecasting for new stores, split stores, new ingredients and recipe changes.
daily forecast accuracy at restaurant level
forecast accuracy at ingredient level
restaurants forecast every day
ingredients forecast daily
stock outs against kitchen waste
indenting from restaurant to supplier
impact of weather, events and promotions on sales
By forecasting every ingredient in every kitchen, every day, Nihilent helped a large quick service restaurant chain order what it will actually sell, cutting waste without putting service at risk.
