Anyone who has followed American Apparel lately knows that they have been having a bit of trouble to say the least. Between losing their cool factor in pop culture, ousting their eccentric founder for scandalous behavior, massive layoffs, and their plummeting sales and stock price, all I have heard as of late about the former fashion darling has been extremely bad news.
But this weekend I came across an intriguing article praising American Apparel’s Chief Digital Officer Thoryn Stephens for exceptional work in data-driven online marketing.
Stephens’ extensive experience in the industry has taught him that the creation of large followings on social media and engagement across internet sites is only as valuable as said engagement drives the bottom line. To illustrate this point, Stephens talks about a past experience where he paid Kim Kardashian to tweet about the company he was working with. While Kim generated thousands of hits to the company’s website, it drove just 30 orders of an average volume of $30 for the company.
So how can marketers be sure that their digital efforts create and capture value for their firms? Stephens’ suggests the following strategy:
Understanding Customer Preferences and Insights
American Apparel (AA) has prioritized tracking as many of their customers as possible through conversion events such as when they click on AA ads across the web, open an AA newsletter, watch an online video, engage with the company on social media, or buy something from the online store.
The company has been able to achieve this level of tracking by dropping cookies on users’ web browsers, collecting IP addresses, and maintaining robust tracking tools on all their owned channels like Facebook, Twitter, and American Apparel.com.
AA then uses this raw data to constantly run AB tests to determine customer insights and preferences. For example, Stephens and team can run two similar ads at the same time to understand if there is a statistically significant difference in impressions, clicks, or conversion rates; if one ad is deemed more effective than the other, Stephens can test the effective ad against another permutation to see if he can be even more effective, thus learning about what customers are more likely to engage with along the way. This AB testing can be done across all consumer touch points, including AA’s own website to optimize their digital efforts.
Delivering a Custom Experience
Now that AA has run thousands of AB tests and collected results along the way, their data scientists are able to run cluster analyses to determine different customer value segments. These segments can then be targeted more precisely with customized marketing as opposed to relying on a one-size-fits-all advertising approach.
For example, AA has used its cluster data to build custom audiences in Facebook based on demographic data like gender, age, and location to deliver ads on the platform that are more relevant to specific users. Stephens refers to this strategy as “one-to-one” marketing.
Predicting Customer Behavior Based on Past Behavior
Perhaps the most interesting part of Stephens’ strategy for AA digital marketing is using the data he’s collected to predict customer behavior. He is likely able to achieve this using the customer value segments as a guide for not just how and when a given user will want to be communicated with, but also how they will prefer to shop, what items they will want to buy, and how much they are likely to spend.
Armed with this type of information, AA can customize the experience on AmericanApparel.com at the user level to present the right product mix with the right visuals to entice a purchase.
But Will It Work?
This is the outstanding question that I have for American Apparel. Will data-driven marketing be enough to increase sales of the fledgling company? At its core, AA is a fashion businesses with little expertise in data science. Bringing in Stephens was certainly a step in the right direction, but is it too much to expect his team to drive impactful increases in sales when what they’re selling is an extremely tarnished brand image?