Every Online Purchase Begins With One Question
A woman is sitting in her living room scrolling through an online fashion site. She has already fallen in love with a blazer. The styling is sophisticated, the color is exactly what she has been looking for, the reviews are encouraging, and she can already imagine wearing it to an important meeting the following week.
Her finger hovers over the Buy Now button. Then she stops.
Not because she has changed her mind about the blazer. She stops because she still cannot answer one simple question: How will this actually look on me?
Will the shoulders overwhelm her frame? Will the length be flattering? Will the fabric drape the way she hopes? Will she feel as confident wearing it as the model appears in the photograph?
That hesitation may last only a few seconds, but it represents one of the fashion industry’s most persistent and expensive problems. Some shoppers abandon the cart. Others order two or three sizes, planning from the beginning to return what does not work. Still others decide that online apparel shopping feels too uncertain and postpone the purchase altogether.
Different decisions. The same underlying problem: a lack of confidence.
After more than thirty-five years in fashion – as a merchant, executive, consultant, and strategist – I have come to believe that the industry’s next major competitive advantage will not be artificial intelligence itself. It will be the confidence that intelligent technology can help create.
The AI Conversation Is Missing the Human Decision
Artificial intelligence now dominates nearly every conversation about fashion’s future. We discuss AI-powered trend forecasting, automated merchandising, personalized marketing, generative design, inventory optimization, and increasingly sophisticated virtual try-on experiences.
These developments matter. Yet technology has never transformed an industry simply because it was new. It changes an industry when it solves a stubborn human or business problem.
E-commerce did not succeed because websites were exciting. It succeeded because consumers wanted greater access and convenience. Mobile commerce did not become essential because smartphones were impressive. It became essential because people expected to shop wherever they were. AI will be judged by the same standard: does it help people make better decisions?
In fashion, one of the most consequential decisions occurs in the seconds before a customer commits to a purchase. The customer is not evaluating an algorithm. She is evaluating herself in the product. She is deciding whether the image on the screen can become a believable version of her own life.
That is why I believe the conversation should begin with confidence, not technology. AI may power the experience, but confidence is what the customer feels – and what the retailer ultimately needs.
Why TryStyle Entered the Story Early
When I was first introduced to TryStyle, I assumed I was about to see another virtual try-on platform. Virtual try-on has been discussed for years, and many companies have made meaningful progress in helping shoppers visualize products online.
What changed my thinking was not simply the visual effect. It was the larger question the company was trying to answer: How can we help consumers shop with greater confidence while giving brands better information to create, merchandise, and sell stronger products?
At its simplest, TryStyle is a photo-based virtual try-on and Fit Intelligence platform for apparel e-commerce. A shopper uploads a full-length image and key measurements. The technology then creates a realistic representation that allows her to see how a garment appears on her own body rather than relying exclusively on a model with different proportions.
That consumer-facing experience is important because it helps replace imagination with evaluation. Instead of asking, ‘Can I picture this on myself?’ the shopper can begin asking more useful questions: ‘Do I like the proportion? Does the length work for me? Is this how I want to present myself?’
But the visualization is only the front door. The deeper opportunity lies in what brands can learn from the interaction: where shoppers hesitate, which products inspire confidence, how body proportions affect preference, which sizes are considered, and where repeated patterns may point to product or assortment issues.
That is where virtual try-on begins to evolve from an e-commerce feature into something more strategically valuable. I call that larger idea Fit Intelligence.
Great Fashion Still Begins With Great Product
I want to be clear about one point because it matters deeply to me: technology does not replace the product, and it should never diminish the work of the people who create it.
Great fashion begins with exceptional product. Behind every successful garment are designers who turn inspiration into a coherent collection; merchants who understand the customer and the market; technical designers and pattern makers who refine fit and construction; sourcing teams who balance quality, timing, and cost; and production partners who translate hundreds of decisions into something a customer can actually wear.
A garment that appears effortless is rarely effortless to make. Fabric, proportion, color, trim, construction, price, timing, fit, and brand identity must all work together. The customer’s final reaction may be emotional, but the work behind that reaction is disciplined and exacting.
The digital marketplace has not made product less important. It has made the translation of product more difficult. In a store, a customer can touch the fabric, see the color in real light, hold the garment against her body, and enter a fitting room. Online, much of that sensory and personal information disappears.
The challenge, therefore, is not to use technology as a substitute for great product. It is to help customers experience great product with enough clarity and confidence to make a decision. Fit Intelligence should serve the work of designers, merchants, and product teams – not compete with it.
What Thirty-Five Years in Merchandising Taught Me
Fashion has always relied on a combination of evidence and instinct. We studied selling reports, returns, store feedback, customer comments, competitive assortments, and market trends. We also relied on experienced merchants and buyers who could sense when a proportion was wrong, when a color would resonate, or when a product had the elusive quality that makes someone say, ‘I have to have that.’
That instinct remains essential. Fashion is not mathematics. It is part science, part art, and part understanding people.
But much of the information available to merchants has traditionally arrived after the fact. We learned after the season had begun, after returns accumulated, after markdowns became necessary, or after customers explained what did not work. The information was useful, but it was retrospective.
Fit sessions are a perfect example. Designers, merchants, technical designers, and production teams can spend hours adjusting a garment by fractions of an inch. That work is extraordinarily important, but the fit model represents one body. Customer reviews and return reasons later expand the picture, yet they often arrive after the inventory commitment has already been made.
The opportunity now is to create an earlier and more continuous feedback loop. If brands can understand not only what was purchased but also where customers hesitated, what they tried to visualize, and which body-product combinations created confidence or doubt, then the merchant’s instinct is not replaced. It is sharpened.
Defining Fit Intelligence
I define Fit Intelligence as the combination of realistic visualization, body and fit information, shopper behavior, and product feedback that helps consumers make more confident decisions while helping brands improve decisions before, during, and after a sale.
The definition matters because it moves the conversation beyond a single on-screen image. Visualization is the consumer benefit she can immediately understand. Intelligence is the longer-term business capability that can influence merchandising, product development, fit consistency, assortment planning, and customer loyalty.
For the consumer, Fit Intelligence can reduce uncertainty. For the retailer, it can reveal how customers interact with products before purchase. For the merchant, it can add another layer of insight to assortment decisions. For designers and technical teams, it may expose repeated fit or proportion concerns earlier. For executives, it creates the possibility of connectingcustomer experience with product strategy in a way that is more immediate than traditional post-season analysis.
This is why I do not see Fit Intelligence as a replacement for fit expertise or merchant judgment. I see it as a bridge between the craft of creating the product and the customer’s ability to believe in that product online.
Case Study 1: Warby Parker Made Trying Part of the Brand Experience
Warby Parker offers an instructive example from eyewear. The company did not treat virtual try-on as a novelty disconnected from its brand. It integrated digital visualization into a broader ‘ways to try’ philosophy that also includes stores and home try-on.
On its current website, Warby Parker describes its virtual try-on as a lifelike way for customers to see how frames look and fit using a computer, tablet, or phone. The important strategic lesson is not that eyewear and apparel are identical – they are not. The lesson is that visualization is most powerful when it supports an existing customer promise: convenience, reduced uncertainty, and a more enjoyable decision process.
Warby Parker’s approach reinforces a point fashion companies sometimes miss. Technology becomes credible when it feels like a natural extension of the brand experience. Customers do not need to admire the technology. They need to trust the decision it helps them make.
Case Study 2: Walmart Framed Virtual Try-On Around Confidence and Inclusion
Walmart’s investment in Zeekit shows how a large retailer can view virtual try-on as a strategic shopping capability rather than an isolated feature. When Walmart introduced Zeekit technology in 2022, it acknowledged one of the central frustrations of online apparel shopping: understanding how an item will actually look on the customer.
Walmart first enabled shoppers to select from models representing a range of body types, and later introduced ‘Be Your Own Model,’ allowing customers to use an image of themselves. The company explicitly connected the experience to inclusion, personalization, and shopping confidence.
This matters because confidence is not created by presenting one idealized body. It is created by helping more customers recognize themselves in the shopping experience. Walmart’s scale also demonstrates that virtual visualization is moving from experimentation toward an expected part of digital retail infrastructure.
Case Study 3: Google Is Expanding Visualization Across the Shopping Journey
Google’s work in virtual try-on offers another view of the category’s evolution. In 2023, Google introduced generative AI virtual try-on for apparel using models with different body shapes and sizes, with attention to details such as drape, folds, cling, stretch, and wrinkles. In 2025, it expanded the experience so shoppers could upload their own photos for apparel try-on across product listings.
Google’s role is different from that of a fashion brand or retailer. It sits earlier in the discovery journey, where customers may be comparing products across many sellers. That makes its expansion significant: visualization is no longer confined to a single product page or a single brand’s site. It is becoming part of how consumers search, discover, compare, and evaluate fashion online.
The broader lesson is that shopper expectations will not be shaped only by a retailer’s direct competitors. They will be shaped by the best digital experience customers encounter anywhere. As tools become more personal and more widely available, the question for brands will shift from ‘Should we offer this?’ to ‘What strategic value can we create from it?’
While companies like Walmart and Google have made significant advances in virtual try-on technology, I believe the next evolution goes beyond helping consumers visualize clothing—it helps brands make better decisions.
Walmart’s Zeekit and Google’s Virtual Try-On are designed to answer an important consumer question: “What will this look like on me?” By making online shopping more visual and engaging, they help reduce uncertainty before a purchase.
What attracted me to TryStyle is that it approaches the challenge from a broader perspective.
TryStyle does more, it enables shoppers to upload a full-length photograph of themselves and see a highly realistic representation of how a garment looks on their own body as well as how the fabric falls and moves with body. Rather than imagining how an item might fit based on a model with different proportions, consumers can view clothing in a way that feels much closer to looking in a fitting-room mirror. That level of realism can create greater confidence in the purchase decision.
But the innovation doesn’t stop there.
Every interaction has the potential to generate Fit Intelligence—valuable insights into fit preferences, purchasing behavior, and consumer confidence. Those insights can help retailers better understand their customers, support merchants in making stronger assortment decisions, and provide designers with earlier feedback that can inform future product development.
In other words, while the realistic virtual try-on enhances the customer experience, Fit Intelligence has the potential to improve the business itself. That’s what makes TryStyle compelling to me. The realistic visualization is what shoppers experience; the intelligence generated behind it is what can help brands create better products, reduce returns, strengthen customer confidence, and make smarter merchandising decisions.
While companies like Walmart and Google have made significant advances in virtual try-on technology, I believe the next evolution goes beyond helping consumers visualize clothing—it helps brands make better decisions.
Walmart’s Zeekit and Google’s Virtual Try-On are designed to answer an important consumer question: “What will this look like on me?” By making online shopping more visual and engaging, they help reduce uncertainty before a purchase.
TryStyle enables shoppers to upload a full-length photograph of themselves and see a highly realistic representation of how a garment looks on their own body. Rather than imagining how an item might fit based on a model with different proportions, consumers can view clothing in a way that feels much closer to looking in a fitting-room mirror. That level of realism can create greater confidence in the purchase decision.
But the innovation doesn’t stop there.
Every interaction has the potential to generate Fit Intelligence—valuable insights into fit preferences, purchasing behavior, and consumer confidence. Those insights can help retailers better understand their customers, support merchants in making stronger assortment decisions, and provide designers with earlier feedback that can inform future product development.
In other words, while the realistic virtual try-on enhances the customer experience, Fit Intelligence has the potential to improve the business itself. That’s what makes TryStyle compelling to me. The realistic visualization is what shoppers experience; the intelligence generated behind it is what can help brands create better products, reduce returns, strengthen customer confidence, and make smarter merchandising decisions.
Returns Reveal the Cost of Uncertainty
The business stakes are substantial. The National Retail Federation and Happy Returns projected that total retail returns would reach $890 billion in 2024, equal to 16.9 percent of annual sales. In its 2025 returns research, NRF also reported that close to two-thirds of consumers admitted to at least one costly returns behavior, including bracketing.
Those figures cover retail broadly and should not be interpreted as proof that virtual try-on alone will solve returns. Returns have many causes: quality, damage, late delivery, changing customer preferences, misleading product information, and fit among them.
But the scale of returns makes one point undeniable: uncertainty is expensive. Every avoidable mismatch between what the shopper expected and what arrived creates costs across shipping, labor, inventory, markdowns, customer service, and customer trust.
The goal should not be to eliminate returns at the expense of service. A fair return policy is an important part of customer confidence. The more strategic goal is to improve the quality of the original decision so that fewer purchases begin with uncertainty.
Confidence Is a Business Metric, Even When It Begins as an Emotion
Confidence sounds emotional, but its consequences are measurable. A confident shopper is more likely to complete a purchase, less likely to order multiple sizes simply as insurance, and more likely to return to a retailer that consistently helps her make successful choices.
Confidence also affects the quality of customer data. A purchase made with uncertainty tells a retailer less than a purchase supported by meaningful interaction. When a shopper visualizes several products, considers particular silhouettes, compares proportions, and then chooses one, the path to purchase can reveal preference signals that a final transaction alone cannot show.
This is where the most forward-looking opportunity lies. The image on the screen may improve conversion today. The patterns behind thousands of interactions may improve product decisions tomorrow.
That is also why brands must treat the data responsibly. Body information and personal images are sensitive. Trust, transparency, privacy, consent, and clear limits on data use are not peripheral issues; they are essential to the credibility of any Fit Intelligence platform. Confidence cannot be created through a system the customer does not trust.
Fit Intelligence Sharpens Instinct Rather Than Replacing It
The fashion industry sometimes frames technology and human expertise as opposing forces. I believe that is the wrong choice.
The best merchants will continue to have instinct. The best designers will continue to imagine what customers have not yet asked for. The best technical teams will continue to understand fit in ways that cannot be reduced to a dashboard. The strongest leaders will continue to make decisions that require judgment, courage, and experience.
Fit Intelligence can strengthen those capabilities by giving talented people earlier and richer information. It can help a merchant test an assumption, help a designer see repeated customer concerns, help an e-commerce team understand hesitation, and help a CEO connect customer experience with product performance.
Data without expertise can produce shallow conclusions. Expertise without current information can miss changing realities. The opportunity is to combine them.
A Final Thought
When people ask me what the future of fashion looks like, they often expect me to talk about artificial intelligence. Instead, I find myself talking about people.
The woman wondering whether a blazer will flatter her before she clicks Buy. The designer hoping customers experience a garment exactly as intended. The merchant trying to build a stronger assortment for next season. The technical designer working to create consistency across bodies and sizes. The executive balancing growth, profitability, and customer loyalty.
Technology matters because it can serve all of them.
After more than thirty-five years in this industry, I remain convinced that great fashion will always begin with exceptional product, talented people, and a deep understanding of the customer. What is changing is our ability to connect those elements more intelligently and earlier in the decision process.
For decades, fashion companies competed on product, price, distribution, and brand. Those pillars remain. But the next generation of leaders will compete on something equally important: the confidence they inspire before, during, and after every purchase.
AI may power that transformation. Fit Intelligence may organize and extend it. But confidence is what the customer will remember.
And in an industry built on helping people express who they are, confidence may become the most valuable fit of all.
Sources and Further Reading
The case studies and retail-return figures in this article are based on the following primary sources. Interpretive conclusions are the author’s own.
1. National Retail Federation and Happy Returns. “NRF and Happy Returns Report: 2024 Retail Returns to Total $890 Billion.” December 5, 2024. https://nrf.com/media-center/press-releases/nrf-and-happy-returns-report-2024-retail-returns-total-890-billion
2. National Retail Federation. “Consumers Expected to Return Nearly $850 Billion in Merchandise in 2025.” October 15, 2025. https://nrf.com/media-center/press-releases/consumers-expected-to-return-nearly-850-billion-in-merchandise-in-2025
3. Warby Parker. “Ways To Try.” accessed July 2026. https://www.warbyparker.com/ways-to-try
4. Walmart Corporate. “Walmart Launches Zeekit Virtual Fitting Room Technology.” March 2, 2022. https://corporate.walmart.com/news/2022/03/02/walmart-launches-zeekit-virtual-fitting-room-technology
5. Walmart Corporate. “Walmart Levels Up Virtual Try-On for Apparel With Be Your Own Model Experience.” September 15, 2022. https://corporate.walmart.com/news/2022/09/15/walmart-levels-up-virtual-try-on-for-apparel-with-be-your-own-model-experience
6. Google. “Google Introduces New AI Virtual Try-On Feature.” June 14, 2023. https://blog.google/products-and-platforms/products/shopping/ai-virtual-try-on-google-shopping/
7. Google. “Google Shopping: How to Virtually Try On Clothes Using Your Own Image.” May 20, 2025; updated November 20, 2025. https://blog.google/products-and-platforms/products/shopping/how-to-use-google-shopping-try-it-on/
8. Google. “Google’s Generative AI Is Improving Virtual Fitting Rooms.” June 14, 2023. https://blog.google/products-and-platforms/products/shopping/virtual-try-on-google-generative-ai/
THE FUTURE OF FASHION | Maria Pesin / maria@vibeconsulting.co
