American Eagle Outfitters: Well-designed experiments help retailer get measurable ROI

With more pressure than ever to be relevant to changing customer preferences, retailers are continuously innovating with investments in pricing, promotion, assortment, technology, and in-store experiences. Executives need confidence in the ROI of these activities to justify such associated spending.

Over the past two decades, American Eagle Outfitters has used data as the foundation to inform and enhance every customer experience. The company’s data-driven culture, along with the fact that it has always owned product fulfillment from material sourcing through sales and service, has enabled it to remain quick and agile.

“We’ve been on the cutting edge with respect to fabrics, fashion, and overall product-line innovation,” says Jimmy Hunkele, Director of Data Analytics at American Eagle. “Every year, we are making significant investments in marketing, merchandising, and in-store, and we need to make sure that we are making the right decisions on this digital frontier; powerful solutions for analyzing large volumes of data with advanced analytics are critical to achieving this outcome.”

American Eagle chose Google Cloud and partner Accenture to further transform its approach to data analytics and enhance decision-making for different investments across the enterprise.

Migrating data into a singular system

American Eagle was increasingly seeing data from a number of sources that just a few years ago was unimaginable. Web, mobile, and third-party data combined with transaction, inventory, weather, mobility, and other market data presented new opportunities to understand customer behavior and shape it.

The American Eagle team decided to overhaul its approach to data warehousing to increase speed and scalability.

“We saw the value of migrating to BigQuery, Google Cloud’s enterprise data warehouse, to better access and analyze large amounts of information at speed,” says Jimmy. “Our goal was to bring in all of our clickstream data coming from billions of hits to our websites and mobile platforms to better understand web user activity in real time. BigQuery gave us the scalability and processing power to analyze massive datasets that were previously too hard to manage in our old systems.”

American Eagle also took advantage of the native integrations across Google products to fast-track insights on their data. By exporting Google Analytics 360 data into BigQuery, American Eagle was able to better understand its different customer segments on the web and customer targeting.

The results: reduced ETL times, further data consolidation, and optimized running of web analytics and testing platforms through a single system.

Automating data science for better results at scale

American Eagle prioritized the testing platform as a key need and decided to work with partner Accenture. “Retailers live in causally dense environments where lots of things are happening at the same time,” says Wanda Shive, Analytics Innovation Principal Director at Accenture. “This makes it very difficult to attribute what investment is causing the shift in customer behavior, or increased popularity of a product, or greater success at a store. Carefully designed experiments are the only way to truly understand what actions are likely to provide the best ROI.”

“Testing is critical at American Eagle because we constantly evolve our strategies related to new styles, pricing, in-store and digital experiences, loyalty programs, and more,” explains Jimmy.

Experiments demand test and control groups, and American Eagle needed the ability to control for attributes such as trends, merchandise mix, percent of sales on promotions, basket compositions, and more. Once the test is done, it also needs the ability to understand which metrics were impacted by the test on a granular level.

The experiment solution proved to be the perfect choice, as it delivered business interfaces, embedded analytics tools, and reporting features all integrated within American Eagle’s BigQuery environment. With tools including Google Kubernetes Engine and Cloud SQL underpinning the deployment and experts at Accenture working on the project, the new testing solution was up and running in four months. While BigQuery provides the horsepower for the data crunching, Cloud SQL serves as the database for the application to provide the latency for the user.

American Eagle now leverages automated analytics to run in-store tests to optimize pricing, layout, and store location decisions, as well as to improve overall brand experiences based on more dynamic, timely customer insights.

For example, American Eagle can quickly conduct in-store testing on promotions using control data from times before the promotion begins and real-time sales performance. This can tell the retailer whether a buy-one, get-one sale will be more profitable than a buy-one, get-one half-off sale, and exactly what types of customers each of them are attracting and how the program can be further fine-tuned.

The experiment solution built on top of BigQuery opens the doors to testing products more efficiently before a season leading to more successful product placement in stores, and much more.

Building on success with advanced technologies and strategies

Through its enhanced forecasting and real-time insights into customer behavior, American Eagle is better positioned to make stronger decisions impacting product production, supply chain operations, in-store staffing, and more.

“Google Cloud and Accenture enable us to make smart decisions quickly, which is now more essential than ever to success in retail,” says Jimmy. “We’re looking forward to further exploring opportunities using machine learning and other technologies. We view this as just the beginning of our success with Google Cloud and Accenture.”