Infosys® creates a 360-degree customer view to help leading Chinese appliance manufacturer transition to an internet-centric world.
In China, consumers are rapidly shifting to buying online rather than through brick-and-mortar stores. Traditional manufacturers and retailers face intense pressure to adapt and successfully compete in this Internet-driven business environment. For a leading appliance manufacturer in China, the transition presented major technical challenges. The company needed to create a 360-degree, integrated view of its customers that spanned interactions across all channels, including the company’s brick-and-mortar stores, online sales, social media feedback and customer-service calls. Furthermore, the system had to support significant business growth: the company was adding 200,000 new customers and a million transactions every day, with online sales soaring 2,400% in 2014.
To achieve its goal, the appliance manufacturer turned to Infosys, a strategic technology supplier. Infosys was charged with developing a system that provided a single source of accurate, detailed, up-to-date customer information. Goals included improving customer service, while enabling advanced analytics to help the manufacturer drive growth and sell more sophisticated products. The project’s scale outstripped the capabilities of traditional analytics tools: Infosys needed to integrate 20 TB of structured and unstructured data from more than 20 different sources, including 10 years of customer purchase history, 200 million user records and 430 million transactions.
After assessing competing technologies, Infosys chose to develop the system using HPCC Systems® from LexisNexis® Risk Solutions. This provided a big data platform that could handle demanding, diverse requirements — including advanced analytics for marketing as well as fast response times for customer service queries.
The solution provides:
- Analysis of customer purchasing history, demographics and interactions integrated from all sources, supporting customer segmentation and the creation of targeted, automated marketing campaigns. This includes gauging consumer sentiment through natural language processing of 400,000 social media and online interactions each day.
- High performance and reliability to support sub-second query response times for more than 2000 concurrent call center agents.
- The scalability to support business growth over the next two years, with an expected doubling of the number of data sources and total data volume and a four-fold increase in the number of users.
- The ability to add data in near real time that becomes immediately available to agents and other users.
- Rapid development, including the ability to enable ad-hoc queries by end users.
Technical Targets Achieved
- 200 million customers records
- 20 trillion customer transaction data and interaction data
- 1 million structured transaction delta data everyday
- 200,000 new customers every day
- 400,000 unstructured interaction sessions everyday
- 20 heterogeneous data sources
- Ad-hoc full dimension customer segmentation analysis in 40 minutes
ECL and rich embedded functions cuts down the time-intensive data integration workload.High concurrent data queries
ECL and ROXIE can delivery complex query efficiently and with high performance.Unified platform
Perform transactions, analytics, and search functions in a single well-designed unified platform.Reduce hardware cost
Commodity hardware allows scale out to improve cluster computing capability and storage.Increased IT efficiency
Eliminates all unnecessary ETL processes and provides full-featured data manipulation functions to make managing data and building applications easier.Fast data link technology
Keyed-index, payload-index are provided to accelerate data searching and data linking.Data analysis efficiency
Rich embedded machine learning and statistic functions make data analysis much easier.Rich data processing functions
Rich embedded data manipulate functions make application building easy.