A wardrobe management tool that empowers mindful shopping choices.
Ever find yourself sifting through a packed closet, only to feel like you have "nothing to wear"? You’re not alone—80% of us wear just 20% of our wardrobe.
With overconsumption and clothing waste at an all-time high, my senior capstone team set out to create a solution that brings greater awareness to what people already own, promoting more mindful and sustainable shopping habits.
As the sole designer, I took charge of the entire user experience, from the user interface to the hardware ergonomics and the interaction between the app and the hardware.
Reflect is a clothing recognition system that enables users to seamlessly track their clothing inventory and usage. It consists of a camera, an image recognition model, and a user facing app.
Users mount the Reflect camera onto their everyday mirror, snap a photo, and Reflect sends the photo to their mobile app.
The app stores all of their clothing pieces along with all the insights needed to gain a better understanding of their consumption habits.
My capstone group and I share a passion for clothing and fashion — we take pride in our style and how we choose to express ourselves with clothing. But despite our love for shopping, we are very aware of the negative impacts of fast fashion and overconsumption.
The fashion industry is responsible for 10% of global carbon emissions and is the second-largest consumer of the world’s water supply. Approximately 92 million tons of textile waste are created annually by the fashion industry.
I wanted to understand the reason behind clothing overconsumption. With limited resources for this senior design project, user research was lean and scrappy. Conversations with friends revealed that the root of the issue boiled down to:
1.
The allure of new fashion trends and styles drives frequent purchases to stay current.
2.
Renewing closets and regularly buying new clothes has become the norm, fuelled by a consumer culture that promotes the idea that more is better.
3.
As a result, it’s easy to lose track of what’s already owned, leading to buying items nearly identical to what’s already in the wardrobe.
For our team, it felt intuitive to address this issue with some sort of digital wardrobe management app to help users track their clothes and curb redundant and excessive purchases. A quick search in the App Store revealed countless digital closet apps already available.
However, while many closet organization apps exist, few are designed to help users reduce their purchases. If anything, most encourage more shopping with in-app features and recommendations that push users to buy more.
I started my design process by taking a look at what other closet apps are doing, assessing where their strengths and weaknesses lie.
When looking at existing closet apps, there are two very clear gaps in the market:
1.
None of them track how often a clothing item is worn. Knowing what's in your closet is one things, but knowing what is and is not worth owning is powerful.
2.
They all fail to conveniently integrate into users' daily lives. Logging your clothing items into the apps is tedious and cumbersome.
I wanted our app to track clothes as passively as possible, while providing rich insights to help users better understand their closet and transform their shopping habits.
That's when I stumbled upon this medium post by Andre Nader. Someone tracked the cost per wear of all their clothes for an entire year with sewn on NFC buttons.
While it's quite impractical to expect people to sew NFC buttons onto all of their clothes, I really liked the idea. He could just get dressed, go about his day, while valuable data was quietly collected in the background, helping him make more informed decisions when shopping.
At this point, the team was enthusiastic about taking the closet app to the next level. The challenge was determining how to gather user data effectively, with minimal user effort.
Drawing inspiration from Andre’s side project, I began ideating with no idea too far-fetched.
The smart mirror quickly gained traction within the team. It was highly convenient since users already check their outfits in the mirror daily. This solution would allow users to effortlessly log and track their outfits with minimal setup.
However, the high engineering effort and cost associated with developing a smart mirror were significant concerns. That said, I collaborated with the PM to scope down to a more feasible MVP.
Opting for a camera instead of the smart mirror involved some trade-offs. While the smart mirror would offer an ideal, hands-free interface right in front of users, eliminating the need to use their phones, it came with high costs and complex engineering.
On the other hand, a camera presents a more affordable solution, making it accessible for a wider audience if it were to go to market.
Reflect features a camera that seamlessly mounts on your mirror and connects to an app on your device. Simply snap a photo of your outfit each day, and gain deep insights into your wardrobe, helping you understand your clothing habits and make informed decisions.
In an ideal world, our image recognition model would be perfect, but that’s not yet the case. To account for any inaccuracies, users are prompted to confirm the match.
The final confirmation screen effectively communicates the user's task. By pairing the photo with the clothing description, it clearly highlights the existing item Reflect is matching, ensuring users understand the connection at a glance.
Once outfits are captured, each clothing piece is stored in the user’s "Closet," creating a space to clearly and comprehensively view their wardrobe. I initially thought of replicating a physical closet experience—users would swipe through their clothes on their phone like they would through hangers.
Inspired by one of my favourite movies, Clueless, I wanted it to feel interactive and fun. However, given that most people have between 100 and 150 items, this idea quickly proved impractical, as it would be neither scalable nor user-friendly.
I pivoted to a classic grid layout, which is much better for managing photo-heavy content. This layout is not only visually appealing and organized but also scales effortlessly, making it suitable for any size closet, whether it's just 5 items or over 100.
From the grid, users can click into an individual item, allowing them to view each piece in their wardrobe, complete with individual data points and photos of every time they’ve worn it.
Since the image is the most immediate and recognizable element, the chosen design positions the header underneath to follow the natural top-down scanning pattern, making it easier to identify items. Floating buttons were used to optimize space, ensuring longer clothing names fit seamlessly without cluttering the screen.
As Reflect accumulates data on users' clothing and their usage over time, its true strength lies in presenting this information in a compelling and digestible way.
I started by listing all the types of data we could track and organizing them into categories.
I then presented these metrics to five users, asking them to rate each one based on how compelling they found it.
Using their feedback, I refined the list of metrics to include in Reflect. Inspired by Apple’s Health app, I admired how each card neatly presents a metric, enabling users to tap in for deeper insights.
With that approach, Reflect's Insights page allows users to quickly glance at key metrics like usage, environmental impact, and cost while also providing the option to explore the details if they wish.
Users really liked the Insights page, but many were unclear about what their data meant. One comment struck a chord:
"30% of clothing utilized in the past 6 months doesn’t really tell me much. Is that normal? Too high? What should I be aiming for?”
While wearing everything in our closets within six months is ideal, it’s unrealistic given factors like seasonal items or special occasion outfits. So what is a "good" target metric?
The more I thought about it, the more I realized how many factors influence wardrobe utilization. To provide users with a meaningful target, we first need to know more about their individual circumstances.
The key metric I focused on is clothing utilization—the percentage of your wardrobe that you actually wear. Since this number can vary widely based on personal factors, we needed to gather insights about each user's unique situation.
To address this, I developed a quick onboarding flow with questions to understand their lifestyle, allowing Reflect to set a practical, personalized goal for each user, making the utilization metric more meaningful.
I considered adding more questions to capture lifestyle nuances regarding dress code, uniforms, and athletic apparel. While these questions could refine the utilization metric, they would also lengthen onboarding and complicate the technical implementation. I prioritized keeping the process quick and straightforward, saving further personalization for future iterations.
Once a target metric was set, the next step was finding tasteful moments to encourage users to work toward it. I aimed for a tone that’s clear, informative, and optimistic, ensuring users feel motivated and empowered rather than nagged.
Here’s how to get started: Set up the app, mount your Reflect camera on your mirror, turn it on, and set the 5-second timer. Step back, strike a pose, and let it capture our outfit!
Soon after, Reflect sends a notification to your device. The message varies depending on whether Reflect identifies the item as new or existing.
Track how often you wear each item with Reflect. If Reflect hasn’t recognized an item before, it will add it to your closet as a new entry. But when Reflect does recognize something you've worn before, it matches it to the existing entry.
To ensure accuracy, simply confirm the identified item. With each confirmation or correction, Reflect's iterative learning model improves, becoming smarter and more precise over time.
If Reflect makes a mistake, simply correct it by choosing the right match from the suggested options. These options are ordered from the most to least likely match based on Reflect's predictions.
Your digital closet gives you a clear, comprehensive view of your entire wardrobe, ensuring nothing gets lost in the dusty back corners.
Easily search, filter, and sort to find exactly what you're looking for.
There are times when you might be in a hurry and forget to check themselves in the mirror, or maybe you don't want to invest in a camera setup. Manually add clothing pieces using your camera or upload photos from your camera roll.
Discover how your shopping habits affect both the environment and your budget. Explore trends and gain insights into your personal style to shop more effectively and sustainably.
If you're not completely sold on how these Insights would actually get users to rethink their clothing consumption, here are some specific use cases to provide tangible instances where Reflect can really make a real impact.
You’ve got a favourite white t-shirt you wear all the time, but have 14 others and somehow keep buying more. Reflect steps in with its "Repeat items" feature, showing you how many similar shirts you already own and nudging you with a reminder: "Do you really need that 16th white t-shirt?"
You see clothing trends come in and out like clockwork thanks to fast fashion. Without even knowing it, you fall victim to these trends and now find yourself with a bunch of pieces you don't like and regret buying. Over time, Reflect helps you realize which items are your go-to favourites and which ones were impulse buys that barely see the light of day. With this insight, you can make more thoughtful decisions when shopping, focusing on what you’ll actually wear.
There is so much more the team and I want to achieve with Reflect. While the camera MVP is a solid start, it falls short of our goal to passively and seamlessly track wears.
Right now, the camera MVP has a couple pain points:
1.
Users need to manually turn the camera on and off.
2.
Taking a photo requires pressing a 5-second timer button.
3.
Users can’t see if they’re in the camera’s frame while the photo is being taken.
4.
Clothing confirmation has to be done through the app.
5.
There's no way to bulk add items at the start, so users either need to manually add their entire wardrobe or use Reflect for about six months to gather meaningful data.
Our ultimate goal is to create an experience where users simply get dressed, glance in the mirror like they always do, go about their day, and receive meaningful insights effortlessly, served up on a silver platter.
Finally, we want to turn Reflect into a chrome plug in. With the wealth of data Reflect collects, we can influence users' online shopping habits in a meaningful way. The extension could analyze the contents of users' shopping carts and provide personalized recommendations on what to keep or avoid, helping users make smarter, more sustainable choices.
What happens when a user realizes they have a closet full of unworn clothes? Most people would probably just toss them. A natural next step for Reflect would be to close the loop by integrating second-hand marketplaces or donation centres into the app. Clothes stay out of landfills, someone in your network gets a cute new item they'll appreciate, and you earn a few extra dollars.
One of the biggest challenges of this project was learning to effectively scope down and determine what was feasible given our circumstances. This process taught me the importance of engaging in cross-functional conversations with team members from different domains—hardware, software, and product. By understanding the tradeoffs each domain presented, I was able to make informed decisions that balanced ambition with practicality.
This project was a blast because we were all seriously into the problem space—clothes, fashion, and making a positive environmental impact. Working on something we genuinely cared about made it super fulfilling. Plus, teaming up with my closest friends meant we were not only getting stuff done but having a ton of fun along the way. It was the perfect mix of passion, creativity, and good vibes—I couldn’t have been happier with the whole experience
EmailLinkedin