We partnered with one of the biggest food
chains in India to deploy a data-driven system
to introduce value-based offerings

INDUSTRY OVERVIEW

Combos are a great way to boost upsell and cross-sell opportunities for restaurants. While combos might seem like an easy route to cash in on increasing consumer wallet sizes, the lack of balanced and systematically identified combos could drive customers across the aisle.

THE PROBLEM

One of the biggest challenges facing restaurants is how to increase revenue and customer footfall. One way to do this is to introduce value-based offerings, such as combos. However, identifying the right set of combos can be a challenge, especially for large chains with multiple outlets.

SOLUTION

We implemented AWS Glue to automate the ETL process. AWS Glue is a serverless service that makes it easy to extract, transform, and load data from a variety of sources. We also used AWS Glue to create a single ETL job that could be used to extract data from all of their data sources and load it into the data warehouse.

METHODOLOGY

We worked closely with the store managers to map out customer preferences and buying patterns. Our advanced Machine learning algorithms helped identify combos that could drive value for the customer, and at the same time ensure profitability for the brand. Customised metrics were designed to zero in on these combos line-ups. We designed the right offering which also accounted for the channel of operation, number of items to be bundled and number of pax. per ticket.

Our recommendations were incorporated across all 30+ stores in India.

HOW DID WE ENABLE CONSUMPTION?

AWS Glue was used to design multiple value-based combo variants and promotional offers. This was done by using AWS Glue to gather data about customer preferences and buying patterns. This data was then used to create a variety of different combo variants and promotional offers. AWS Glue was used to test customer adoption of the different combo variants and promotional offers. This was done by creating a variety of different experiments and tracking the results of those experiments. The results of these experiments were then used to determine which combo variants and promotional offers were the most popular. The best received combo variants and promotional offers were scaled pan-India as permanent menu combos. This was done by using AWS Glue to deploy the combo variants and promotional offers to all of the stores in India.

OUR IMPACT

Our ML algorithms delivered a 22% increase in revenue through direct cross-sell gains, thus increasing average ticket size. Test stores also saw an increase in customer footfall by 5%. .