AI+Human is magic: 3 Lessons from the Stitch Fix IPO About Personalization

Do you remember when was the last time you had fun shopping for new clothes? Me neither.

Over the past several years, the cycle of searching for, trying on, buying, and returning clothes has lost whatever joy it used to bring. In retail shops, you spend more time trying to find parking or standing in the checkout lines, than actually looking for what you like. Online, the sheer amount of selection is dizzying; you’re caught in a maze of price comparisons and unreliable customer reviews. And when you finally make a purchase, it’s inevitably the wrong size, triggering an arduous return process.

So, when I started using Stitch Fix recently, it was like stumbling upon an oasis of personalization, usability, and — yes! — fun in the otherwise lifeless fashion-buying desert. If you have read some of my earlier posts, you know by now that I am a fan. And I’m not alone. Founded by entrepreneur Katrina Lake in 2011, Stitch Fix filed for its IPO last Thursday. According to its filing, the startup has grown to serve more than 2 million customers, the vast majority of which are repeat buyers. In fiscal 2017, Stitch Fix reached nearly $1 billion in sales.

With so many other clothing retailers struggling, and in the face of intense competition from giants like Amazon, how is Stitch Fix succeeding?

In a word: personalization. On the road to its IPO, Stitch Fix has demonstrated three crucial lessons about the power of 1:1 marketing, made scalable by technology.

1. Every Customer Is Different; Every Experience Should Be Different

Here’s how StitchFix works:

You fill out a fashion profile on their website and pay a $20 styling fee. Using its proprietary data science/AI fashion-matching technology, plus the expertise of a personal stylist, Stitch Fix selects a mix of five clothing items and accessories and ships them to you. You can try on the clothes, purchase what you want, and send back the rest. Shipping is free both ways. Customers like me appreciate Stitch Fix for its seamless buying experience. But we love it for its personalization.

  • No two shipments are the same. Each is tailored to the fashion taste of an individual customer, yet dialed into modern trends.
  • As you continue to use Stitch Fix, it “learns” more about you. The fashion matching technology hits home more often than not, and your returns become less frequent.
2. Data Can Make Your Customers Feel Human

One of the worst aspects of digitization is being treated as faceless. The genius of Stitch Fix is that it uses numbers to re-humanize people. Their personalization technology starts by gathering 85 data points on each of its customers. The Stitch Fix IPO filing proclaims: “Our data science capabilities fuel our business.” Stitch Fix Chief Analytics Officer Eric Colson heads up the algorithm team that matches customers to clothing. He previously did a similar job at Netflix, another company that struck gold after realizing its role wasn’t to push products, but to engage with customers on an individual level — and smart use of data science is the way to scale the whole thing.

3. AI + Humans = Magic

The proprietary Stitch Fix algorithms are powerful tools. And they work best in the hands of experts who know how to deploy them to solve real-world fashion dilemmas. When I ordered from Stitch Fix, a computer program may have done the heavy analytics, but one of the company’s 600 stylists applied the finishing touches. As YEC points out, “In this way, the recommendation technology enables humans to do their jobs better, not the other way around.”

The Stitch Fix IPO emerged from the realization that personalization for each and every customer can be scalable — thanks to data science, algorithms, and artificial intelligence.

Using Real-Time Customer Data for 1:1 Marketing

With the right technology, you can apply the same approach to your marketing. Whether you’re reaching out through email, push notifications, SMS, or any other channel, you can tailor your message to each customer, drawing on real-time data for 1:1 marketing.

To learn more, download “The Path to Predictive 1-to-1 Marketing”.