How American Apparel Is Moving to Value-Driven Data Analytics

Thoryn Stephens, chief digital officer for American Apparel, once worked for a company that paid Kim Kardashian to tweet about its products. With 15 million followers, Kardashian delivered a huge audience. But when Stephens did the math, he found that those 15 million Twitter followers drove 1,200 visits to the company’s website, resulting in 30 orders with an average order volume of $30. His takeaway: “Focus on value metrics.”

Speaking at ClickZ Live San Francisco, Stephens explained how American Apparel is experimenting to find the most meaningful data as it moves up what he calls the “data maturation curve.”

Beginning at the lower end, the data maturation curve includes:

  • Understanding customer preferences and insights
  • Doing a business-impact analysis
  • Predicting customer preferences and business impacts based on historical performance of the customer and other customers in the segment

Stephens advocates a circular process of using data to derive hypotheses and testing the hypotheses, and then refining the insights based on those tests.

Getting to Know Them

Stephens emphasized that consumers at every stage of engagement have value, whether the brand knows them or not. Unknown users may consume or engage with the brand, but the brand has no data on them. They gain the most value when they move to a brand-owned or operated property, at which point they can be cookied in order to begin to build an anonymous profile. At this stage, information can come from geo-location via the IP address, as well as any onsite behavior that is tracked.

Already, Stephens said, the brand can begin to start to tailor the experience to that consumer with ads and dynamic content on websites.

Users become known via conversion events, such as site registration, social login, subscribing to a newsletter, downloading an app, or happiest of all, buying something. At this point, companies can pull in data from third parties in order to further optimize everything about the experience, from ads to social media to email to video.

“From there, you can push variables and create custom audiences through Facebook or other channels to tailor the experience to be one to one,” Stephens told the conference audience. “The greatest amount of value comes from your known users.”

Once you begin to understand your users, you can do cluster analysis to determine different value segments.

What They’re Worth

According to Stephens, the lifetime value of a customer is an important metric, but it’s not necessarily the most revealing. A former executive at Fox Networks, Tillster and Beachbody, he recounted how Fox used Adobe analytics to understand how to drive video ad starts and views of long-form video. It came up with four segments: “Foxaholics,” casual watchers, international viewers and passive users.

Here’s where the data versus insights story got interesting. While passive users seemed to be the least valuable because they didn’t watch video or see ads, further analysis showed that these people used Fox’s schedules to find out what to watch. That data helped justify billions of dollars’ worth of agreements with Fox’s distribution partners.

“This was a major epiphany for the business,” Stephens said. Fox then segmented these users and showed them a home page with a more prominent schedule.

American Apparel Everywhere

Under Stephens, American Apparel is consolidating messaging not only in digital channels, but also in physical stores. For example, the fashion retailer rolled out beacons in some stores for the 2014 holiday season.

The brand has a mobile-optimized website, which Stephens said “has to be square one of your strategy.” It’s developing mobile apps, although he seemed less sure about their value.” Do our consumers need another app? Will they use it – and what type of consumer will?” he asked rhetorically.

Stephens is more interested in Apple Passbook and Apple Pay, saying, “It’s more of a contemporary conversation with your consumers.”

Finally, Stephens emphasized that data is useless if the rest of the company can’t understand and use it. “Data communication has to be part of your overall strategy and a core premise in your data maturation strategy,” he said. He’s found the most effective thing is to sit down face-to-face with marketers and walk them through the data.

The goal, he said, is “forensic storytelling. Take the data and distill it into three bullets that are actionable.”

Related reading

Flat design modern vector illustration concept of website analytics search information.
Website landing page vector graphic
Vector graphic of a laptop surrounded by random items intended to illustrate measurement, such as a crane (which is lifting a video thumbnail into the air), a lightbulb, a pair of compasses, a cog, a ruler and a pencil.