Are the challenges of modern day retail solvable with data science? Personal styling service Stitch Fix thinks so.
The San Francisco, Calif.-based company has forged a new kind of retail business model that uses data and AI to serve curated, personalized fashion boxes to its customers.
Read also: How to Implement AI and Machine Learning
To do so, Stitch Fix employs a team of more than 85 data scientists that oversee a bevy of machine learning algorithms underpinning its operations -- from client styling and logistics to inventory management and product design.
But it's the data that really lets Stitch Fix work its algorithmic magic.
Leveraging data to deliver personalization
Stitch Fix customers first fill out a style profile to form a baseline of their likes and dislikes when it comes to fashion. Stitch Fix then asks for insights and feedback on items customers receive in their curated box, called a Fix, to get an idea how each item lined up with the their style, fit, and price preferences.
The result is a data-driven customer feedback loop that's helping Stitch Fix close the gap between customer information and experience.
"We leverage data science to deliver personalization at scale, which we don't see anywhere else in retail," said Cathy Polinsky, the chief technical officer at Stitch Fix. "Data science is not just part of our culture -- it is our culture and it's woven into every aspect of our business."
While the data is undeniably crucial, Polinsky stressed the yin and yang between humans and machines when it comes to styling. Machines provide the initial filters for stylists by optimizing and conducting rote calculations that would require an immense amount of human time. However, the human stylists are key to understanding the nuances of customer requests and making sure their experiences are personalized.
"At the core of what we do is a unique combination of data science and human judgement. Our human stylists make our algorithms better and our machine learning helps our stylists perform better," Polinsky said. "By combining the art and science of styling we're able to create a far better client experience than anyone else in retail."
Data science and engineering made equal
Internally, Polinsky said the data science and engineering teams are equal on the value chain and closely aligned to encourage collaboration and experimentation. Since 2016, the company's engineering team has doubled in size to 120, and Polinsky said the additional bandwidth has made a tangible impact on the team and the company as a whole.
"One of the amazing aspects of Stitch Fix is that no department is 'king.' While some engineering teams at big tech companies can do no wrong, at Stitch Fix, all functions and teams are equally important and valuable," said Polinsky. "The partnership between engineering and data science is essential to constantly improving our client experience."
Machine learning delivers measurable results
Management of the company's technical talent falls in the purview of Eric Colson, the chief algorithms officer for Stitch Fix. In his role, Colson oversees the company's data science and machine learning projects and works to extend the benefits of machine learning across departments.
"We really rely on the algorithms team for framing problems with mathematical equations," Colson said. "For instance, one of our data scientists tinkered with a genetic algorithm and applied it to apparel to predict what would be a successful piece of clothing that doesn't exist today. We brought that to the merchandise team and now they can use that as a tool."
In terms of ROI, Colson said machine learning has delivered measurable results, including increased revenue, decreased costs, and boosts to customer satisfaction. "It hits the top and bottom lines in very real ways," he said.
Maintaining that competitive edge
Meanwhile, Colson said the draw of the company's machine learning agenda has helped it outcompete for data scientists in cut throat, talent-starved Silicon Valley.
"It's always a tough field to compete in, but we do really well," Colson said. "We have people who've left the field of astrophysics to work in the world of fashion, and the reason is high impact -- they can see what their work is going towards."
Looking long term, Stitch Fix -- which rang up $1 billion in revenue in 2017 -- is confident it'll easily maintain a competitive edge against retail upstarts trying to mimic its data-centric business model.
Colson posits that the corporate structure of Stitch Fix, with its value system framed to support a data science team, is something legacy retailers and new age startups will struggle to replicate.
"Other companies don't have our structure and can't easily copy it," Colson said. "We embraced early on the need to leverage data and anticipate client needs. Companies that fail to do the same will be left behind."
- Enterprise IoT projects: Data, ML, security, and other key factors
- Free PDF: Sensor'd Enterprise: IoT, ML, and big data
- Survey shows that most businesses are taking steps to secure IoT data
- How to use machine learning to accelerate your IoT initiatives
- How to create a data strategy for enterprise IoT
- How to create a security strategy for IoT
- South Korea's IoT in full swing: From water meters to AI-powered smart buildings
- Successful IoT deployment: The Rolls-Royce approach
- Five ways your company can get started implementing AI and ML
- Why AI and machine learning need to be part of your digital transformation plans
- Should Amazon be your AI and machine learning platform?
- Should Google be your AI and machine learning platform?
- Should Microsoft be your AI and machine learning platform?
- Should IBM be your AI and machine learning platform?