In the competitive beauty products retail market, companies like Ulta Beauty must constantly evaluate their strategies to stay ahead and meet customer demands. Ulta Beauty, with more than 1,300 stores in the U.S. and a strong focus on customer service, continually seeks ways to expand its market presence.
In 2018, Ulta recognized the need to upgrade its critical systems, including its aging e-commerce system, and transition to the cloud.
“As we evolved, technology improved, and social media started to become an important part of any e-commerce platform, we realized that we needed to find ways to react more quickly to customer needs,” said Omar Koncobo, IT director for e-commerce and digital systems at Ulta Beauty.
Over time, Ulta had made improvements by transforming its Oracle-based e-commerce platform into Oracle ATG Web Commerce, an application that automates and personalizes the online shopping experience. However, Ulta faced challenges in terms of capability, scalability, and monitoring. During peak times, such as Black Friday, spikes in traffic could crash the system. The physical infrastructure that the system ran in made it difficult to deploy new capacity efficiently. Even if the company chose to add more servers to handle those traffic spikes, it left them with servers they didn’t often need, wasting money.
To address these issues, Ulta’s IT team introduced queuing technology to manage the influx of customers, but this led to frustrations and customers leaving the queue.
Ulta’s IT team also struggled with monitoring error rates and order volumes, which prompted them to implement application performance monitoring tools like Sumo Logic. These tools allowed the IT team to detect and handle problems before system crashes occurred. “If we could see that a crash was on the way and we needed to reduce how many people were coming to the site, we’d get an alert,” Koncobo explained. “That would allow us to reduce how many people we let in.”
As stressful as these technical challenges were, Ulta’s primary concern was how they affected customer experience. The company aimed to integrate social media seamlessly, gain more flexibility in promoting and pricing products, and provide an overall positive and customer-friendly service.
Ulta’s monolithic system architecture posed obstacles to achieving these goals, as even making small changes required modifying the entire system, Koncobo said.
More Than a Cosmetic Upgrade
Ulta made a strategic decision to migrate the entire infrastructure to the cloud and develop a cloud-native platform. The company embraced a microservices and API-first approach, using Google GKE as the engine. Ulta standardized on CommerceTools, a commerce system built with microservices, APIs, and the cloud in mind. This new approach allowed the IT staff to fix or deploy specific services without affecting the rest of the system.
During the transformation, Koncobo carefully selected technologies from the old infrastructure to keep and added new products to enhance the system. New products included Amplience, which allows easy creation and management of web content; XCCommerce for promotion management; and Akamai as the entry point for the entire system.
Although the new system operates alongside Oracle ATG for now, it has already demonstrated significant improvements. The microservices architecture and enhanced accessibility of the new system have made it easier for Ulta’s IT team to address customer issues quickly. For example, the IT team can swiftly adapt to circumstances like shipping delays, Koncobo noted. “We can go into the system and make a quick change to remove the zip code or display a message that notifies customers,” he said. “And we can do it without impacting the homepage, payment systems, or anything else.”
Additionally, with the ability to isolate and monitor each component separately, the new system helps Ulta secure its infrastructure. Furthermore, the development team integrated a continuous integration/continuous development pipeline into the system, ensuring code is scanned at various stages.
The Personalized Future
As Ulta continues to make progress with its system transformation, the company has set its sights on its next project: enhancing personalization options for customers.
Koncobo’s team plans to incorporate the technology of QM Scientific, a company that Ulta acquired for its recommendation engine. That technology will be combined with an AI/ML-based search algorithm from Adeptmind.
“Today, [customers see] the same thing, but our goal is to be one-to-one,” Koncobo explained. “We want everything a customer sees on the site to be tailored just to them based on their preferences.”
About the author
Karen D. Schwartz is a technology and business writer with more than 20 years of experience. She has written on a broad range of technology topics for publications including CIO, InformationWeek, GCN, FCW, FedTech, BizTech, eWeek and Government Executive.
