A mobile app that makes selecting and matching clothes easy for color deficient people.
Design Challenge, UX/UI Design, Accessibility
Picking out outfit in the morning takes time. How can we streamline the process of planning to wear and transform what used to be a chore into something more enjoyable?
While choosing outfit is a tedious task for people with normal vision, it is an extremely frustrating process for color deficient people. Hueman is an app designed to help color deficient people select and match clothes by building up their virtual wardrobes and giving outfit recommendations based on weather, events and individual preferences. Instead of sticking to the basic colors, color deficient people can then enjoy a wide range of outfit options without hassle.
Check out the interactive prototype: here.
Tools: Interviews, Pen & Paper, Sketch, InVision
Duration: 1 week
My user research was structured around two key questions:
- What are the key factors that influence people's decision on what to wear?
- What pain points do people have in the current process of planning what to wear?
To answer these questions, I first made a survey to examine the general patterns on how people plan what to wear. I collected 40 responses from college students and young working professionals. Some of the interesting findings I got from the survey were:
1. People seek for sense of novelty and variety.
"I wish my closet can recommend me new ways of pairing things."
Due to the busy schedule and personal preferences, people tend to wear certain pieces of clothes in their wardrobes way more often than others. They want to find new ways to pair the existing clothes, and more effectively utilize the clothes that they haven't worn in a long time.
2. People find cleaning, organizing and arranging clothes drudgery.
"Please help me organize my closet! It's a mess now."
"I'd make everything magically self-cleaning."
When being asked if they had a magic wand and could change anything in their closet, a tremendous number of people mentioned that they wish their clothes could be more organized and cleaned automatically. These comments suggested that there are great opportunities to improve how people work with their clothes.
3. People who are color deficient need a lot of assistance in buying and selecting clothes.
"Sometimes I take a picture and send it to a friend who can tell me what color it is, and if it's 'my color'."
One of the respondents brought up a really interesting issue that I had never thought of before - the problems that color deficient people face when they are shopping or choosing clothes. He mentioned that when he was at home, his mom would lay out what he need to wear every day. Ever since he moved to college, he tried to stick to the basic colors so that there would not be room for mistakes.
Reframe the Problem
Although all the three insights were fascinating to me, I decided to go with the last one as there is a strong need for color deficient people to be more independent and embrace a wider range of colors in their cloth selections.
Learn about Target Users
The online blog posts and community discussions have revealed the pain points color deficient people have experienced.
- They need to find people in the store, or send pictures to families or friends, to ask for the colors of clothes. They often feel embarrassing and helpless in this process.
- They need others to choose clothes for them in the morning.
- They have developed their own system of labeling clothes. Some use plastic braille labels, and others use different types of stitching patterns to represent different colors and appearances of clothes.
- Lots of color deficient people choose to limit their cloth choice to dark palette, which is supposed to be "safe".
Based on the research, I created a user persona to guide myself through the ideation phase.
I came up with six ideas to address the user's pain points above. I created a storyboard for each idea and then interviewed 3 potential users. Below are the feedbacks I gathered for each idea.
Matching Clothes #1 - Recommendation System
Idea: Based on weather and the personal schedule, the app will recommend Jay what to wear each morning.
- Checking weather before deciding what to wear is already part of people's daily routine, so people don't need to do extra things.
- If the recommendation system is robust, it will consider both color and pattern of clothes.
- Some people don't like to be recommended the specific items. They are more acceptable to color or pattern matching recommendations.
Matching Clothes #2 - AR App
Idea: When Jay goes to the wardrobe in the morning, he holds up the phone and the app shows the color of each cloth in an AR setting.
- This idea will be useful if people already have a certain piece in mind.
- Telling the color solely isn't going to help people pick the clothes. They want to know if the color is appropriate on certain occasions.
- People want to know about the nuances in colors ("I already know it's blue, my challenge is to tell the difference between shades.").
- The lighting might affect the detection of color.
Matching Clothes #3 - Smart Hangers
Idea: The smart hangers sense the color of clothes. When Jay reaches for a cloth, the sensor sends data to his phone which tells him if it matches with the other colors he had selected.
- People will have the control on each step so that they know each piece well.
- The sensor's detection will not be affected by lighting.
- The app gives direct and clear feedbacks on whether different pieces could go together.
- People are concerned about the costs.
Shopping #1 - Volunteer System
Idea: Jay can send a picture of cloth to a volunteer system to ask people to help identify the color.
- It doesn't require too much efforts for color deficient people nor volunteers.
- Knowing the color is not going to influence people's purchase decisions. They are more interested in whether the cloth matches to the rest of his existing clothes.
- It requires sufficient amount of volunteers for the system to work.
Shopping #2 - Shop Assistants Locator
Idea: When having questions, Jay calls shop assistants through the app.
- The idea can potentially be generalized to other customers.
- Will shop assistants always be around? Some people were questioning about the solid needs for this idea.
Shopping #3 - Match with Owned Clothes
Idea: After Jay scans the cloth, the app indicates how it can be matched to other pieces in Jay's closet.
- People felt strongly about this idea! People thought finding the matching pieces from wardrobe would be really handy.
- People liked the sense of independence this app would give them.
What I learnt from the speed-dating
Different from what I expected, knowing the exact color of cloth won't be very helpful for my target users. They're more interested in knowing the match between different pieces. And the match will include more factors than colors, such as texture and patterns.
Color deficiency has such a wide spectrum that only showing the simple color categories won't suffice the target user's needs. As one of my interviewees mentioned, "I can tell it is blue; it's the shades that I struggle the most". Showing the nuances in colors will not only be enabling, but also empowering my users.
Inputting the new items into the virtual wardrobes should be easy to achieve. For many users, this might be the first interaction they will have with the app. If the process is too complex or disappointing, they will soon lose faith in the app.
Learn about users' style preferences when they first onboard
Research Insight: Styling is such a subjective thing that the outfit recommendation should be personalized.
Action: When users first log in the app, they will be given a collection of cloth set that they can either swipe left or right to indicate their preferences.
Result: The data collected here will be used to optimize the outfit match results.
Recommend outfits based on weather and user's schedules
Research Insight: Currently, target users often rely on others to select outfits for them on a daily basis. Weather and schedules are two key factors that help people decide what to wear.
Action: The home page shows the weather and reminds users of important events they have on that day. Users can swipe to view the recommended outfits. When they click on the individual set, they can see the specific items and the colors in more details.
Result: It streamlines user's current flow of checking the weather and schedules before determining the outfit. It also gives users great independent on selecting their own clothes.
Use computer vision to smooth the process of adding new items to virtual wardrobe
Research Insight: The process of adding new items needs to be easy and fast to motivate the users to complete this task.
Action: When clicking on the plus button in the virtual closet page, users will be led to a camera view where they can inspect the colors of cloth and take a picture as well. Computer vision will analyze the image and find the possible match from the database.
Result: It enables users to quickly learn the colors and add newly-purchased items. By ensuring that the virtual wardrobe is updated constantly, it builds the foundation for the daily recommendation feature.
Give users flexibility to add items by searching for brand and style number
Concern: If a cloth's color and pattern are pretty common, there might be too many search results.
Action: Users can search for brand and style number to add the new clothes.
Result: It gives users flexibility in how they want to add the newly-purchased clothes.
Help users evaluate whether the selected items can go together
Research Insight: In addition to the colors, the target users are more interested in knowing how different pieces of clothes can go together.
Action: Users can add items to the match canvas via searching clothes in their closets or taking pictures of new items. The match canvas will evaluate the items and give a match score and the style of outfit.
Result: Users will know how well their selected pieces will match, and how the set of outfits might be appropriate for different occasions.
Enable users to find matching outfit based on selected items
Research Insight: When shopping, the users want to know the colors of clothes interested. They also want to know if they can match with the existing clothes in their wardrobe.
Action: After adding the new item onto the match canvas, users can add specific items from their virtual wardrobe to see how they may go together. They can also click the "Match Items From Closet" button to check out what can be matched from their own closet.
Result: It gives users ideas about how they can wear the cloth they are interested in before even purchasing it.
Why did I choose color deficient people as my target users? Even though lots of the features in my app can be useful for people with normal vision, color deficient people will be more motivated to use the app, especially for adding new items to their wardrobes - a process that doesn't give users immediate gratifications. While it might be a "good-to-have" app for people with normal vision, the Hueman app can truly bring the sense of independence and confidence to color deficient people in their daily lives.
What would I do if I have more time? Due to the time limitations, I didn't have a chance to develop lo-fi prototypes and test out my ideas with the target users. Additionally, since color deficiency has such a wide spectrum, I wish I could talk to people at different places in that spectrum to learn more perspectives.