TellApart Case Study
The retail business is a fierce world of high transactions and low margins. Within this world, e-commerce is particularly competitive; and excellence in online retail requires operational acumen. This is why, when online retailers need an edge against the competition, they turn to technology platforms such as TellApart for assistance.
Amazon is the only platform that allowed us to address the diverse set of technical challenges, which we faced initially, as quickly as we did with as small an engineering team as we had.”
With its suite of marketing tools—a customer data platform, predictive customer analytics, and a next-generation display ad retargeting application—TellApart helps online retailers identify (or “tell apart”) their best customers and prospects. Online retailers can then use this curated data to send personalized marketing messages to their best prospects. In doing so, retailers can drive substantial growth to their bottom-line.
Although data warehousing, predictive analytics, and customized marketing solutions have all been available independently in the past, TellApart believes it is the first company to bring all of these components together in a cloud-based platform. They’ve also introduced to the market a fresh, innovative business model to online retailing—retailers pay TellApart only when shoppers click through TellApart display ads and actually make a purchase.
Why Amazon Web Services
Launched in 2009 and headquartered in Burlingame, California, TellApart has leveraged Amazon Web Services (AWS) from the beginning. From the earliest prototype models to their most current deployments, which serve tens of thousands of queries per second (QPS) to hundreds of millions of end users, TellApart has run every major part of its technology stack through AWS. In fact, the company does not run any processes on physical servers.
Co-Founder & CTO Mark Ayzenshtat esays, “Amazon is the only platform that allowed us to address the diverse set of technical challenges, which we faced initially, as quickly as we did with as small an engineering team as we had.” There were many challenges to be met:
- Serving a large volume and variety of request types with minimal latency
- Serving a large quantity of retailer product images with low latency
- Recording logs for all requests, reliably storing them and collecting them from a central repository for subsequent analysis
- Running a complex batch work-flow of jobs numerous times throughout the day and aggregating the data
- Scaling up to accommodate sudden traffic spikes initiated by high volume shopping days or new retailers coming online with TellApart’s services
Meeting these technical challenges continues to be very important to the company and TellApart’s implementation of Amazon’s products has been instrumental to its success. The extensive scope of TellApart’s operations dictated an AWS solution that comprised a variety of AWS products:
- Amazon Simple Storage Service (Amazon S3) stores accumulated raw server data
- Amazon SimpleDB stores millions of user property records and retailer product records
- Amazon Elastic Block Store (Amazon EBS) stores the logs and provides underlying storage for the HBase cluster
- Amazon Elastic Compute Cloud (Amazon EC2) instances are used in four different availability zones for serving ad impressions, real-time bid requests, and user recall site actions
- A separate cluster of Amazon EC2 instances to run batch data processing, including aggregating user data
- Elastic Load Balancing (ELB) balances requests across TellApart’s front-end server instances
- Amazon CloudWatch and autoscaling monitors the server instances and ensures availability
- Amazon Simple Query Service (Amazon SQS) manages TellApart’s proprietary in-house data replication across regions
The AWS products mesh perfectly with TellApart’s operations to meet all of the goals. Co-Founder & CEO, Josh McFarland, relates the following experience: “We’ve run our production infrastructure on AWS since the earliest days of TellApart. To be honest, there were many folks who said we couldn’t build a real-time bidding engine on top of AWS. We’ve proven them wrong, and today, we respond to tens of thousands of real-time bid requests per second, each in under 120 milliseconds.”
Having dismissed the naysayers, TellApart (and its clients) is thrilled with the outcome. As CTO Mark Ayzenshtat explains, “AWS has enabled us to build an infrastructure that is efficient, elastically scalable, low latency, and redundant. AWS is a key partner of ours, enabling TellApart to offer the fastest and most scalable technology in display ad retargeting to our retail clients.”
Within this world, e-commerce is particularly competitive; and excellence in online retail requires operational acumen. This is why, when online retailers need an edge against the competition, they turn to technology platforms such as TellApart for assistance.
Benefits of AWS
- Responds to tens of thousands of real-time bidding requests per second
- Maintains response rate of 120 milliseconds
- Gained efficiency, scalability, and redundancy
AWS Services Used
Amazon Simple Storage Service (Amazon S3) is an object storage service that offers industry-leading scalability, data availability, security, and performance.
Elastic Load Balancing
Elastic Load Balancing automatically distributes incoming application traffic across multiple targets, such as Amazon EC2 instances, containers, IP addresses, and Lambda functions.
Amazon SimpleDB is a highly available NoSQL data store that offloads the work of database administration.
Amazon CloudWatch is a monitoring and observability service built for DevOps engineers, developers, site reliability engineers (SREs), and IT managers.
Amazon Elastic Block Store (EBS) is an easy to use, high performance block storage service designed for use with Amazon Elastic Compute Cloud (EC2) for both throughput and transaction intensive workloads at any scale.
Amazon CloudFront is a fast content delivery network (CDN) service that securely delivers data, videos, applications, and APIs to customers globally with low latency, high transfer speeds, all within a developer-friendly environment.
Amazon Elastic Compute Cloud (Amazon EC2) is a web service that provides secure, resizable compute capacity in the cloud. It is designed to make web-scale cloud computing easier for developers.
Amazon Simple Queue Service (SQS) is a fully managed message queuing service that enables you to decouple and scale microservices, distributed systems, and serverless applications.
Companies of all sizes across all industries are transforming their businesses every day using AWS. Contact our experts and start your own AWS Cloud journey today.