Retailer Takes a Stance on the Cloud and ERP

Selling apparel and other items online is a tough proposition—even for large retailers. There’s no shortage of competition, and site performance is often a make-or-break proposition.

For Stance, a retailer that established an e-storefront in late 2010, the route to success has passed through an enterprise resource planning (ERP) application in the cloud.

“We entered the market with the goal of infusing some personality into the sock and underwear markets,” says Andrew Spencer, director of technology for Stance. “Exponential growth and periodic heavy traffic volumes have placed enormous demands on the infrastructure and the business.”

The firm, which operates brick-and-mortar stores in New York City and San Francisco, as well as a Website that sells products in 40 countries, offers limited edition products for Star Wars, the National Basketball Association and many other groups and celebrities. When these items become available on the site, huge bursts in traffic result.

For example, when pop star Rihanna offered a limited edition sock at Stance in 2015, traffic exceeded 1,000 requests per second at peak periods. Moreover, the ERP system had to keep up with the demand. “It’s critical to scale the technology and the business to accommodate large variations in traffic,” Spencer points out.

Cloud Supports a Variety of Systems and Functions

Although Stance began using Amazon Web Services in 2013 to deliver a more robust infrastructure and dynamic bandwidth, it migrated to RedisCloud and began using Redis Labs ERP in June 2015. The transactional environment—which operates as a private cloud within the Google Compute Engine (GCE)—supports a variety of systems, including ERP functionality, a shopping cart, business tools and analytics.

Overall, the system accommodates upward of 3,000 different products, and it is able to find a single item from all the stock-keeping units (SKUs) in a matter of milliseconds. It delivers lightning quick searches by storing snapshots of the company’s entire product database in a memory cache. “The results are almost instantaneous,” Spencer says.

Even during peak periods, Stance is able to keep latency below .007 milliseconds by using the in-memory database and auto-scaling technology, he reports. The stock notification system uses Redis to fetch inventory information from the system of record and transmit it to requesting clients.

During special sales events involving limited edition items, such as the Rihanna sock, the site also uses a countdown timer that requires checkout in 20 minutes or items are returned to the inventory pool. “We were able to develop and implement the countdown feature in less than a month,” he explains.

Spencer says that the cloud and ERP technology position the company competitively for the future. It delivers a highly flexible and scalable foundation that supports other use cases for in-memory data stores that are required in a fast transactional retail environment. The cloud-based approach also delivers far more robust security than what Stance could develop on its own, he says.

Moving forward, “We have tremendous resources and expertise at our disposal,” Spencer reports. “We do not have to invest in capital or staffing to ensure that we are operating in the most efficient and effective way possible.”