Zappos: Mastering the Art of Personalized, Data-Driven CX

Add bookmark

The world was stunned and saddened this past weekend to learn of the death of former Zappos CEO and business visionary, Tony Hsieh

One of the very first major success stories of the digital era, during his tenure at Zappos, Mr. Hsieh not only revolutionized the online shopping experience, he also set new standards for customer-centricity and employee experience. While other big retailers built their empires on low prices, selection and convenience, Zappos focused on, “developing relationships, creating personal emotional connections, and delivering high touch ('WOW') customer service." Most famously, Zappos (along with Amazon, who acquired Zappos in 2009) made CX perks like free shipping and 365-day returns the norm for retailers of all types. Zappos also became well known for its unique, “a little weird” corporate culture and it’s flat organizational structure or “holacracy”. 

Though entire books could be, probably will be (and for that matter have been) written about Mr. Hsieh’s immense business and philanthropic accomplishments, we would be remiss if we didn’t take a look at Zappos’ singular approach to innovation. As one of the OGs of the digital age, Zappos has stayed at the forefront of cutting-edge technology embracing everything from social media to warehouse automation to, more recently, AI-powered personalization. 

However, at least under Mr. Hsieh’s tutelage, Zappos seemed to not only understand but embrace a concept so many other organizations never fully grasp. That the true value of technology does not lie in its capacity to replace people, but rather its ability to mitigate our innate deficiencies, elevate the human experience and increase interpersonal connection. 

 

Zappos Personalization Engine 

There’s a lot of talk out there about personalized product recommendations. The standard way of doing this, in a nutshell, is to apply AI algorithms to a customer’s past purchase history to more or less predict what they might purchase next. However, the problem with this approach is that, as the system ends up only suggesting similar items to what a person has already bought, it’s not exactly accurate. 

To remedy this, the data science team at Zappos has taken things a step further by focusing on customer intent. For example, let’s say you just bought a pair of winter boots. Instead of showing you other winter boots or “similar items,” Zappos personalized recommendation engine will show you complimentary products such as winter coats, mittens or wool scarves. 

The Zappos team also noticed that the number one reason people returned items was poor sizing. In fact, many people were purchasing multiple pairs of the same shoes, trying them on, then returning the extra ones that didn’t fit. For a company that offers free, 365-day returns, these trends could be very costly. 

Leveraging Zappos’ massive amounts of shoe size data, Zappos was able to develop an ML-powered sizing prediction model that recommends which shoe size a customer should purchase. According to Ameen Kazerouni, Zappos’ Head of AI/ML Research & Platforms, this predictive model was so accurate it “drastically cut down our sizing related return rate without compromising the customer experience. It's a great example of AI being a win for the org and a win for the customer experience at the same time.”

In the time since Zappos introduced it’s shoe sizing predictive model, it’s also expanded this capability to accurately predict apparel sizes, eventually collaborating with Amazon SageMaker.  

 

Semantic Search

Traditional, lexical-based search engines, in a nutshell, match the user’s search keywords with the products in the store. However, relying solely on the words a person enters simply wasn’t cutting it for the customer-obsessed folks at Zappos. Afterall, who hasn’t struggled to translate a search into succinct keywords. 

Given its massive inventory, Zappos wanted to minimize the number of unsuitable items a customer has to sort through in order to identify their desired product. So they put together a team of about 20 data scientists and software engineers to build a search engine that understands context and, once again, customer intent. 

As Ameen Kazerouni told VentureBeat, “We realized that English is a very funny language in the sense that many, many words are heavily overloaded. They have many different meanings depending on the phrase that they occur in,” said Kazerouni. “So the first thing we set out to do was understanding search terms, taking in search terms, and looking at customer behavior and building machine learning models that could create what are known as word embeddings.”

Dubbed “genetic algorithms,” Zappos new AI-enabled model not only takes into account user keywords, but also that individual’s purchase history and general fashion trends as well.

For example, take the phrase “classic short.” This phrase could refer to shorts or a popular style of Ugg boot. However, using a system of neural networks, the search engine is able to accurately predict if that person is looking for shorts or new boots based on that user’s real-time and past shopping history as well as whatever other data Zappos may have on them.

As a result, it significantly reduced the number of search mishaps and time wasted on running repeated searches. Zappos has also achieved higher search-to-product-clickthrough rates and raised the position of customer selections in search results.

 

A More Human Approach to Innovation

There’s no doubt about it: Zappos is a forward-thinking, technology-driven company. What makes them so successful though is not the tools themselves or the size of their IT budget, it’s the understanding that innovation isn’t about technology but humanity. They don’t develop new technology solely to replace human workers, reduce costs or even increase competitive edge, but rather to enhance human experience both within the organization and beyond.

Case in point, though Zappos probably has the ability to use chatbots on their site, they don’t. They prefer to communicate with customers over the phone. However, behind the scenes, call center reps use AI/ML assist bots to help them answer questions, find products faster and deliver the best service possible. This is a perfect example of how companies can leverage cutting-edge technology strategically to boost human connection and stay true to their values.

 


RECOMMENDED