Retail & Consumer Products
Headquarters: New Delhi, India
- Reduced TAT for critical processes like Sales Plan from 7 days to 1 day
- Month end closing has been reduced from 3 days to 1 day
- Efficiency of sales force has increased by 40%
- Machine learning algorithms have contributed to 5% increase in sales
Godfrey Phillips India Limited (GPIL), flagship company of Modi Enterprises, and one of the largest FMCG companies in India has been producing some of the most popular cigarette brands in the country. GPIL also manufactures and distributes iconic brand Marlboro under a license agreement with Philip Morris. They have also expanded business interests in chewing products, mouth fresheners, confectionary and retail.
“We learned that data was scattered across multiple source streams. Different teams and functions used disparate sets of data for the same reporting purposes. This could potentially lead to anomalies in datasets used for analysis and also cause data reconciliation issues. We realized that we needed to consolidate all the data into a centralized repository,” said Farman Khalid, Head – BI & Analytics, Godfrey Phillips India Ltd.
GPIL wanted to integrate its data across the enterprise and create a single source of truth from which various internal stakeholders and decision-makers could easily access new consumer demand signals, leverage new insights, enable real-time monitoring of sales, develop advanced analytics and make fact-based decisions. GPIL also wanted to enable integrated data management services across the enterprise so that teams could effortlessly create analytics crucial to the operations of the sales program.
A data lake implementation was envisioned in the initial phase which could be used as a centralized repository of both structured and unstructured data from disparate applications and external sources.
GPIL realised the need for real-time insights to drive decision-making and inform next steps for the sales team. The company was looking for a cloud solution considering it adopts an agile methodology across the enterprise and scalability of the solution was a priority. Cloudera’s enterprise platform provides the user with the flexibility to choose the environment where the project can be hosted, and also permits an increase or decrease in the number of nodes based on the project requirements.
GPIL created an enterprise-wide data lake using a cloud aggregation and alerting framework leveraging solution from Cloudera. “To meet scalability and flexibility requirements and support the execution of such projects, we created this setup on AWS and engaged with Cloudera for an agile and open source framework that provides the user with the flexibility of choosing an environment based on their specific business need.” Khalid added.
GPIL uses Cloudera’s enterprise platform to create a centralized repository of multi-structured data across disparate applications and external sources. The platform is easy-to-use, enables data science and advanced analytic capabilities leveraging an open-source architecture, rapidly updates predefined KPIs, reduces load on transactional systems and provides visibility using targeted dashboards.
With the help of Cloudera Manager, GPIL can centrally operate the data management tasks and other managed services. Using the Cloudera platform, the company has been able to automate the installation process giving them a cluster-wide, real-time view of hosts and services running.
GPIL has been able to create an enterprise-wide data repository that serves as a single source of insights. Teams across the enterprise can benefit from smart visualizations and data analytics based on a robust and open source framework. On the ground workers now have access to an enhanced and extended insights platform for quick and effective decision making using smart visuals.
Teams effectively and efficiently use big data analysis for predictive and prescriptive data analytics to drive sales and optimize operations such as daily exit reports, supply chain visibility, inventory optimization and computer assisted ordering. There are no longer instances of data inconsistency between systems and they’ve built an internal data science capability using the Hadoop data lake to run the algorithms.
GPIL has been able to improve operational efficiency and reduce manual intervention across the enterprise by standardizing models, forms, and templates across the organization.