This interview was originally published in The Interline’s AI Report 2025. To read other opinion pieces, exclusive editorials, and detailed profiles and interviews with key vendors, download the full AI Report 2025 completely free of charge and ungated.
For fourteen years, personal styling service Stitch Fix has had machine learning at the heart of a mission that ‘s arguably more relevant today than ever: pairing people with products and brands they actually love. With the wider fashion industry now rushing to use AI to help bridge a growing gulf between demand signals and supply, Stitch Fix’s tech-native journey has a lot to teach – and CPO / CTO Tony Bacos has been thinking deeply about what it means to build a business on top of algorithms that give people superpowers, rather than acting as a crutch.
The Interline: Stitch Fix is one of a small cohort of fashion businesses that were built on a foundation of data science and algorithms. This gives you a unique perspective on what’s happening across industry today, because Stitch Fix was extremely early to a data-centric approach to fashion that now seems pretty prescient – and is now something the wider industry is actively pursuing. What has the company learned from pioneering a business model that relies on cutting edge technology? And what have you learned from proving out that model over close to fifteen years?
Tony Bacos: To take a step back, Stitch Fix was founded 14 years ago to solve a common yet deeply personal problem: how do I find clothes that help me look and feel my best?
What we’ve learned since disrupting the retail space over a decade ago is that the consumer pain point we were solving for when we started is at least as relevant today as it was then…and arguably even more so. In fact, our research shows that nearly 90% of consumers experience stress or self-consciousness when trying to find clothes that fit. IBM also released research that shows less than 10% of people are satisfied with in-store shopping and only 14% are satisfied with online shopping.
People want and deserve a better way to shop. That’s what we offer. We make it possible for people to find the perfect styles – and fit – without spending hours browsing stores or sifting through seemingly endless choices online.
We can do this because of the strength of our model. We help people discover the styles they will love by pairing expert human Stylists with best-in-class AI and recommendation algorithms, and by leveraging our assortment of exclusive, well-known and emerging brands.
This year, we have begun to reimagine our client experience and create a more modern and dynamic Stitch Fix. The improvements we’ve made are resonating with clients and we’re seeing that reflected in the momentum in our business and great feedback from clients.
Since the beginning, we’ve known technology alone isn’t enough to deliver an incredible, personal experience. One of the main things we’ve learned over the past 14 years is that the magic happens when tech enhances (not replaces) the human experience. Our model works because it honors both precision and empathy. Trust is earned not only through accuracy, but through the feeling that someone gets you. We are constantly learning and are excited to see how we can continue to evolve to serve our clients.
To consumers, Stitch Fix is a personalised styling service with a human touch; you don’t position it as an AI company. How do you decide when and how to talk about AI? Do you see it – and technology in general – belonging behind the scenes, and quietly powering incredible customer experiences, or is public awareness and adoption of AI changing the conversation?
When communicating to our clients about how we are able to deliver such a personalized retail experience, we focus on some of the biggest shopping challenges today – from the long lines, lack of inventory, dirty and disheveled dressing rooms to the seemingly endless scroll online – and how we address them.
From the beginning, our priority has been creating an experience that is deeply personal, and that starts with the people who serve our clients – our Stylists. We believe that AI is a key lever to improve our client and Stylist experience. We have never leveraged AI for AI’s sake…for us, AI is about help, not hype.
That said, we also believe we have the strongest recommendation algorithms in the industry. Thanks to our continuous feedback loop and billions of data points on style and fit, we’re able to offer personalization at scale, serving millions of clients with a depth of insight that is tailored to each individual. This combination of human understanding and technological precision is what makes our model unique.
Ultimately, we see AI not as a replacement for the human element, but as a way to enhance it. It allows us to give Stylists superpowers by arming them with the tools that help them give clients a more personalized experience. “Quietly powering” is a good way to put it. We’re making smart use of AI as one of many important enablers for the great work that our Stylists do.
In this new era of AI, the conversation is evolving. As AI becomes more visible and widely adopted, we think it’s important for companies to be intentional and transparent about how and why they use it. Customers want to know when they’re engaging with a human versus interacting with an AI-powered experience.
In the last couple of years, Stitch Fix has been undergoing a transformation that’s designed to reimagine both frontend experience and backend operations. Before we get into the technical elements of how that philosophy shows up in the service and the systems behind it, it’s important to call out the emphasis you’re placing on people as the heart of this transformation – which isn’t something that can be said for every AI-adjacent transformation and automation strategy. Why was it so important to keep human beings in focus as you’re changing and futureproofing a very technology-centric business? And how is that philosophy showing up in things like stylist profiles?
AI in combination with the expertise of our Stylists is what enables us to deliver a truly personalized experience that meets every client’s unique style needs. Stylists have been central to the Stitch Fix experience since the beginning. Our community of Stylists is what sets us apart – their expertise and empathy is not something that can be artificially replicated. They are trained to partner with clients on their style journey – to get to know their style and evolving needs. So we can help them look and feel their best as they evolve.
As part of our broader effort to enhance our client experience, we are putting a greater spotlight on our Stylists and their work. Last year, we introduced a new feature called Stylist Profiles that enables clients to get to know their Stylist better. Accessible via the digital Style Card clients receive with each Fix, Stylist Profiles are customized by Stylists themselves and feature background information (think first name photo and areas of expertise), fashion tastes and preferences and styling badges and client testimonials. We’re excited about how Stylist Profiles will inspire richer dialogue between our clients and Stylists and enable us to deliver an even better client experience.
The response from clients has been really positive, and this is just the beginning. We are exploring additional ways for clients and Stylists to engage in more “real-time” connections. We recently shared that the percentage of clients requesting the same Stylist for their next Fix hit the highest level in nearly five years – a testament to the trust our clients place in our Stylists and one of the reasons why our strategy is resonating.
Let’s talk a little bit about how AI and automation support those human stylists. There’s a lot of talk in fashion at the moment about augmenting people’s capabilities, and about using AI to free them up to focus on higher-order tasks. What does that look like within Stitch Fix? Are you using AI, for example, to surface insights for each consumer so that stylists are able to start their creative work with the right foundation, instead of needing to hunt through systems to find a customer’s previous fixes and feedback?
We leverage generative AI to help amplify our Stylists’ talents and unleash their creativity, especially when it comes to finding efficiencies for labor-intensive or repetitive tasks. One example of this is the notes Stylists write to clients that accompany each Fix. Each note includes personalized details about the items included in the Fix and some thoughts on how to style them. This is also where Stylists can answer specific questions sent in their Fix request.
We use OpenAI’s latest models to generate suggested content for the Stylist notes based on the items chosen for their clients as well as the client’s history with Stitch Fix. Think of it as the first draft. AI has a deep memory and leveraging it for note drafting serves Stylists with key details – such as writing the names of the products and recalling past purchases – so that the Stylist can edit and add additional context. It enables us to take the laborious part of the note writing so our Stylists can focus on what they do best: understanding and fulfilling nuanced style requests and building meaningful client relationships.
Another example is our recommendation algorithm (CTSM, or “Client Time Series Model”), which is powered by natural language processing and deep learning. Our in-house recommendation algorithms use client feedback to identify items that match each client’s fit and style preferences. We can process billions of words from the written feedback shared by clients to efficiently and effectively provide curated style recommendations for our clients. This allows our Stylists to focus on curating personalized selections for clients, rather than manually sorting through thousands of inventory pieces.
Another area that decision-makers in fashion talk about obtaining real value from AI is in shortening the distance between obtaining information and being able to act on it. As part of your transformation initiative, you’ve had the goal to take consumer – stylist relationships, which historically involved back and forth communication with an inbuilt delay that made them feel perhaps more like penpal letters than chats, and to bring them closer to real-time. How did you approach redesigning that onboarding and interaction experience? And what role has AI played in it?
Our transformation is grounded in fully realizing the potential of the model that we introduced 14 years ago. It’s all about delivering on our mission – to help people discover the styles they will love that fit perfectly so they always look and feel their best.
To do this, we took a critical look at our client experience and began to reimagine how we could best serve our clients today. We have four areas of focus: First, creating a more dynamic way for our clients to engage with us, from the initial sign up to our regular connection points. Second, making the Stylist more visible and creating opportunities to deepen relationships with clients. Third, building more flexibility and engagement into the experience so clients have more ways to shop with us. And fourth, offering a stronger assortment with more fresh and new styles.
The improvements we’ve made to our experience are resonating with clients and we’re seeing that reflected in the momentum in our business. We have some exciting work planned as it relates to Gen AI and style visualization. We believe we have the opportunity to help our clients better understand what our styles can look like and provide inspiration within our core experience. There is exciting work to come!
Thinking a little more deeply about interactions between AI and internal systems, how does Stitch Fix approach things like inventory planning and merchandising – making sure you have the right brands, the right sizes and so on – and where have you either considered employing AI or already deployed it to streamline those processes?
We have developed a proprietary AI-driven inventory management tool that analyzes our exclusive transactional data and client feedback to predict demand at both the individual style and client level. This empowers our merchandising team to make smarter, more efficient buying decisions across key categories. With it, we’re able to drive more relevant inventory buys, improve margins, and increase client satisfaction as a result.
How do you see AI influencing or supporting the future of Stitch Fix? It might be unifying different elements, such as the Fix model and your curated eCommerce experience, or it could be something completely new.
We are investing in inspiration in all its forms. We recognize that inspiration looks different for everyone, and we are committed to offering the most personalized and client-centric shopping experience.
With AI, we’re not just streamlining choices, we’re sparking imagination. Our goal is to help clients see themselves in new ways. Everything we do is about helping our clients find the styles they will love that fit perfectly so they can be more confident in what they wear. I believe AI can help with that.
By combining the power of AI with the insight and empathy of our Stylists, we offer something truly special. Together, they create an experience that’s deeply personal to each Stitch Fix client.
