Skip to Main Content

Quick Service Restaurant

How a Restaurant Accelerated Cluster Segmentation and Personalization

Google cloud-based data analytics.

A large quick service restaurant chain wanted to implement a cost-effective analytics platform to extract greater value from the vast amount of data it captures through its point-of-sale staffed registers and in-store ordering kiosks, mobile app, and delivery service.

The Imperative for Change

The company and Publicis Sapient collaborated to develop a Google Cloud Platform (GCP)-based solution that uses machine learning and custom-developed five ML algorithms to better understand and predict customer behavior and preferences.

The Transformative Solution

The company, Publicis Sapient, and Google collaborated to develop a Customer Data and Analytics Solution that imports data from the company’s data stores into a Google Cloud Platform analytics hub for ingestion, storage, processing, and visualization. Data is processed through BigQuery; then Google Cloud ML machine learning models are applied, generating data insights and predictions based on five algorithms requested by the company:

• Two descriptive algorithms:

            o Recency, frequency, and per-ticket spending

            o Product preference

• Three predictive algorithms:

            o  Customer churn

            o  Purchase propensity

            o  Lifetime customer value

The Business Impact

Creation of the first pilot in the Japan market took approximately one month, processing a year’s worth of first-party transaction data. Production began immediately afterward, and the solution was deployed in-market in Japan at the end of 2019. Initial results show new insights previously unachievable with the more manual processes used in the larger regions; for example, one analysis of loyalty program members predicted that getting customers who visit twice yearly to come in once more during a year could generate as much as $35 million in additional revenue for the region.

 

 

Rapid Development
From pilot to production and fast test-and-team cycles in implementation
Flexible Importing
of disparate data sets to accommodate the unique needs of individual market regions, enhancing the marketing architecture they already have in place
Sharonyka Kumar
Sharonyka Kumar
Group Vice President, Global Head of Google Practice at Publicis Sapient