Real-time analytics has the power to transform businesses offering actionable insights that enable faster, precise and more effective decisions. But the effectiveness of real-time analytics is largely dependent on the availability of data. To draw business insights using analytics, companies require a structured decision process with predefined logic and real-time access to data.
Let me put this in perspective.
Ecommerce is currently at the forefront of a technology-led disruption. Real-time analytics holds immense potential for the sector and empowers retail organizations to understand and serve their customers more effectively with real-time insights on customer behavior and changing market realities.
Ness works closely with an e-commerce company that serves more than 19 million customers in 150 countries. It has a strong international presence with 25 localized sites throughout the globe. Its websites have approximately one million unique visits daily generating approximately 32 million clicks every day.
High abandoned cart rate, low conversion rate and poor rate of returning customers were the most common business challenges faced by this organization. Any undue delay in analyzing data was impacting its decision-making ability and could result in business losses.
Another challenge usually faced by decision makers was having to deal with a fixed set of reports with fixed visualization. They depended on the technical teams every time there was a need to customize the reports.
To stay competitive, it was extremely important that the company provide a better user experiences through personalization to improve conversions and control the abandoned cart rate. And the ability to do that depended immensely on the availability of user data. Data needed to be analyzed when it was collected from the users’ browsing habits for better personalization and recommendations.
But the entire process of collection and analysis of data, presentation of reports and taking relevant business decisions was so time-consuming that often it was no longer relevant to the user anymore.
The solution for real-time decision making
As part of the solution for our client, Ness created and used ANDruid, a real-time analytics and reporting framework that supports handling of billions of events in real time and visualizations of high dimensional data sets. With ANDruid we ensure the data is fresh, the analysis is faster and business decisions are swift. This enables the decision makers to provide real-time personalization, to better understand the customers and make decisions based on facts by providing smarter, meaningful and efficient solutions.
ANDruid is built with Open Source tools like Kafka, Druid, Cassandra and D3.JS. This framework applies user clicks analysis to understand behavior and patterns, provides in-memory and lock-free ingestion and has fast aggregation, highly scalable/available, parallelized architecture which supports thousands of concurrent business users.
In this solution, user click events are injected into Kafka using Node.js service and then written to Druid platform. Real time and historical data is stored into Cassandra after writing meta data information to MySQL.
Queries are served by the Druid platform to the requester who in turn can present the data either in the dashboards/websites. These real-time analysis reports can aide the business users to make faster and effective decisions.
By deploying ANDruid, the business team can now view reports in real time with a huge range of customization built with D3. It also provides a vast set of visualization effects and tools that can be built by the decision makers. They can build the reports the way they want, not limiting themselves to just the predefined templates, enabling better business insights for faster and easier decision making.
Ness continues to work with this client empowering them to take better decisions and harness the value of data to attain real business results.