India’s e-commerce market is projected to reach $200 billion by 2026 due to increased internet penetration, analytics, and transactions. First impressions and compelling imagery are key to capturing visitors’ attention, as is a deep understanding of the consumer shopping journey. To succeed, brands need to create online experiences that complement in-store shopping. 

Tata CLiQ—a platform from a Tata Group business venture, Tata Unistore—was launched in 2016 as an online marketplace for brand-loving consumers. The marketplace connects customers with over 1,200 local and international brands across categories including fashion, electronics, and footwear. The platform also launched its premium and luxury lifestyle destination, Tata CLiQ Luxury, to cater to consumers who focus on quality and craftmanship. In addition to an online marketplace, the business is also a direct seller for selected merchandise categories. 

To provide an immersive, omnichannel shopping experience, Tata CLiQ combines the convenience and service of shopping in stores with the range and speed of online shopping. To do this, the business has championed what it calls “phygital shopping”, with in-store pickup (PIQ), alterations, returns, and exchanges available at over 200 CLIQ and PIQ centers across the country. “Our focus is not only on conversions. We want to bring back the joy of shopping in the online world,” says Dharmarajan K, chief business officer for beauty at Tata CLiQ.

Seeking to Eliminate Downtime and Boost Efficiency

During its initial years of operation, Tata CLiQ used a private cloud to run its infrastructure. However, reliability and scalability were major challenges, which led to persistent inefficiencies. Internal teams had to manually monitor infrastructure 24/7 to quickly remediate system issues. During peak sales periods, they often had to scale up infrastructure three to four times higher than what the business projected to avoid losing sales or risking damage to the company’s reputation. Downtime—both planned for scaling and unplanned due to system overload—was frequent. 

To boost efficiency, Tata CLiQ decided to migrate to the Amazon Web Services (AWS) Cloud. The decision was based on a compelling proposal from the AWS team as well as Tata CLiQ’s leadership team’s experience with AWS. The company worked with AWS Partner Tata Communications Limited (TCL) on planning and implementation and appointed TCL as its managed services partner.

Migrating Oracle RAC and SAP hybris to AWS

Tata CLiQ’s SAP enterprise resource planning (ERP) applications were among the first workloads migrated to AWS. For its customer care operations, the company uses SAP ECC, which is integrated with the Tata CLiQ marketplace. The operation also relies on SAP hybris as a mission-critical component. 

Before migrating to AWS, Tata CLiQ faced teething challenges in ensuring sufficient capacity of its SAP hybris nodes to handle peak traffic during sale periods. Its Oracle Real Application Cluster (RAC), which stores more than 30 TB of data, was also subject to frequent downtime. 

Now the business uses Amazon Relational Database Service (Amazon RDS) for Oracle and SAP on AWS, both of which feature automation tooling to scale on demand. Dharmarajan shares, “In the six months since we migrated to AWS, I can’t recall any database downtime. Efficiency has increased and we no longer face performance-related infrastructure issues. ”

Scaling on Demand with Cost Efficiencies

After resolving its stability challenges, Tata CLiQ began a cost optimization journey on the AWS Cloud. AWS led deep-dive sessions to identify idle and underutilized resources, followed by right-sizing across all resources including database instances, Amazon Elastic Compute Cloud (Amazon EC2) instances, and Amazon Elastic Block Store (Amazon EBS) volumes.

As a result, the company’s overall infrastructure cost decreased by 30 percent in the first three months. The business has also benefited from continued cost optimization and detailed visibility on AWS. “We are now meeting our budgeted cost figures. This used to be a big concern, as our actual versus budgeted costs would vary plus or minus 20 percent,” Dharmarajan says. “Predicting cost is absolutely critical for us to expand profitably at this stage.”

Tata CLiQ has been growing 100 percent annually in the past two years. On-demand scaling has likewise been invaluable for its growing business. Traffic levels can vary widely on Tata CLiQ, rising 500 percent one month and dropping 10 percent the next. “In such scenarios, the flexibility to continually adjust infrastructure up and down is critical to maintain uptime as well as manage costs,” Dharmarajan explains.

Gaining Transparent, More Secure System Operations

Visibility into security and system operations has also improved since migrating to AWS. TCL and AWS guided a detailed security baseline review of Tata CLiQ’s environment, recommending changes to access controls and logging, in addition to enhanced tracking metrics. 

The business is now using Amazon GuardDuty to continuously monitor for malicious activity and unauthorized behavior. In addition, Tata CLiQ leverages services including Amazon CloudWatch and AWS CloudTrail to collect detailed logs that track user actions and enhance traceability when an error occurs. “This is the highest level of explainability and granular visibility we’ve ever had,” says Dharmarajan. “We’re finally able to pinpoint what, when, and where something went wrong so we can learn from errors and prevent recurrences.”

Tailoring Communications Based on Data-Driven Analytics

By removing the need to manually monitor security and instances during scaling, Tata CLiQ has empowered its teams to innovate and enhance site features. One example is the development of customized product description pages for high-value items such as refrigerators. When buying big-ticket items, Dharmarajan explains, consumers typically seek out more information before purchasing. 

The business is also benefiting from data consolidation and native integration among data services on AWS, with a data lake in Amazon Simple Storage Service (Amazon S3), a data warehouse in Amazon Redshift, and AWS Glue for serverless data processing. Dharmarajan says, “Our ability to slice and dice all the data in our system on a near-real-time basis has dramatically improved on AWS.”