Facilitating
Decisions

Simple - A Digital Shoe Rack App

Role

Product Designer, Product Manager

Responsibilities

Prioritization, Visual Identity, User Testing, UX Design, Prototyping, Animated Mockup

Teammates

Andrew Lou, Cassie Zhang, Gibson Chu, Karishma Ahuja

Duration

12 Weeks

Overview

Simple is a digital shoe rack that helps you make better decisions from picking, organizing, to purchasing shoes. Considering weathers, occasions, and personal preferences, Simple recommends the best shoe for the day. This product was developed by a group of product management students at the University of California, Berkeley in the spring of 2019.

View Final Result
The Problem

Limited cognitive bandwidth to utilize of the closet

“Does it spark joy?”
— Marie Kondo, the host of “Tidying Up with Marie Kondo”

Indicating by the popular Netflix show, “Tidying Up with Marie Kondo,” the problem of cleaning unused items at home resonates with a large amount of audience. From clothing, books, to Komono (kitchen, bathroom, garage, and miscellaneous), countless items are left unorganized in every house in the show. An article by The Guardian points out that “most purchases are not fully rational.”

“We often buy dresses for the people we’d like to be, and hold on to things because they remind us of good times. But a new survey has revealed that UK shoppers own £10bn worth of clothes they do not wear.”
“Why are our wardrobes full of unworn clothes? Because most purchases are not rational”, The Guardians

In addition to reducing impulsive purchase, Marie Kondo suggests that another way of utilizing people’s closets to be more aware of what they already have. With information overload in the current society, people only have so much cognitive bandwidth to deal with their cluttered closet when they are getting dressed. Many people have a regular rotation that they wear weeks after weeks. Therefore, a good portion of clothing is hidden under the massive pile of clothes.

How might we design a product that represents users’ entire collections and facilitates their daily outfit decisions?
Target Customers

Busy people in need for clothing advise

The main target customers are busy people who have too many clothes or people who need recommendations for outfit combinations. The typical quality of these two personas is that they both don’t have much time to spend on outfit decisions. Thus, Simple is here to help.

Competitive Analysis

Identifying the clothing market opportunities

To differentiate our product from existing solutions in the market, we analyzed several competitors. Stitch Fix, for example, focuses on expanding one’s closet through tailored recommendations. We also researched on daily outfit suggestions like Amazon Echo Look or Cladwell.

Market Position and Differentiators

Combining the strengths of our major competitors, we would like the market position of Simple to be exceptional at automation while maintaining recommendation accuracy. We take weather, colors, activity level, and the occasion of the day into consideration. We believe that we could deliver users more suitable daily recommendations when we consider more external factors than the other services. The user will also be able to set preference from our recommendations for the algorithms to train the model to make more accurate predictions in the future.

External factors for more accurate predictions

Weather

Color

Activity Level

Occasion

Design Principles

Seamlessly integrating into a part of life

Solving an issue in people’s daily life, we would like to design the product to be integrated without being an extra layer in people’s morning routines. Moving forward from the research insights to the design execution, we utilized the following principles throughout the product development.

  • Effortless and direct interactions

    Making outfit decisions from daily recommendations should be the forefront of the product. Furthermore, the process of reconstructing people’s closet in their mobile devices should be as effortless as possible.

  • Facilitated informed decision making

    We have to make users trust the daily outfit recommendation to make their decision efficiently. Providing the factors from the algorithm in a minimal yet sufficient manner would be the ideal situation for users to utilize this tool in their busy morning routines.

  • Organized and digestible collections

    Given the number of clothing items a user possesses, the digital closet would have to be filtered in a way that clearly shows what the user owns. Moreover, the visual identity of the product should be colorless because all the different colors from different clothing items could clutter the collection easily.

Ideation

“How might we” and “Crazy 8”

To rebuild the digital closet and suggest daily outfit, we sketched out explorations for “how might we let users log their clothing item effortlessly” and “how might we provide scannable yet informative recommendation” collectively.

My sketches for item-logging.

My sketches for daily recommendations.

Risks

Complexity in digital closets and clothing recommendations

However, several issues like the complexity of clothing type came up during the brainstorming discussion. When researching existing product in the “smart closet” market, we also discovered that the complexity of clothing collection and combination could be a significant risk for our product.

  • Much more complicated than just tops and bottoms

    The combination of the daily outfit could be extremely complicated. There are a massive amount of different styles, functions, brads, materials, etc. for clothes. Since it is nearly impossible to collect all kinds of clothes into the database, users would have to log all their clothing items into the digital closet manually. Also, the number of clothing combination would be a significant obstacle for delivering the recommendation system in our vision.

  • Too many items to log in the app

    The users probably own an extensive collection that is difficult to browse through on a day to day basis to switch to a digital closet service. However, users would be required to log all their clothing items into the app to maximize the value of the app. The amount of possible time and effort to rebuild their closets in mobile devices might not be as cost-effective as we thought it would be.

  • Hard to provide immediate value to the user

    Because clothing style is subjective, it would be difficult for the users to receive immediate appreciation when they use the service. Moreover, there would be a predefined style from the user’s clothing collection. Thus, adopting Simple might not result in a drastic increase in social recognition.

Scoping Down

Piloting through footwear

After getting feedback from the instructor, we decided to focus on shoes for the remaining time of the product development to validate our hypothesis:

  • Recommendation accuracy

    We want to see if having these external factors — weather, occasion, and activity level — in addition to personal factors, help improve outfit recommendations people want, thus increasing user engagement.

  • People’s willingness to build their digital closet

    We want to test out a template-based upload process to facilitate an easy-to-use user closet registration, thus increasing initial user retention.

We choose shoes because people can’t complete their daily outfit without putting their shoes on. Also, imagine wearing a pair of non-water proof shoes in a pouring day or wearing high heels for a long walking day. The outcomes of inaccurate shoe decision are relatively measurable comparing to the whole outfit.

An essential part of the out with distinct functions and apparent pain points

  1. A necessary part of the outfit
  2. Less variety in terms of styles, functionality, and categories (compare to the entire outfit)
  3. A large enough and growing market to sustain the product
  4. Shoes are less adjustable during the day (heels, rainy day)
  5. Mostly visible on a shoe rack or in the closet to reduce the physical obstacle
User Testing

Evaluating the initial prototype

Building from the clothing concepts from our ideation, we created a low-fidelity prototype. We then conducted some user testings with four people with three qualifications: a) owning a significant number of shoes; b) frequently wearing different shoes; and c) experience with using fashion services.

Initial User Flow
User Feedback

Not valuable enough as a minimum viable product

  • Interested in purchasing recommendations

    All users acknowledged that the daily recommendation could be valuable to solve how they neglect other shoes that aren’t in their core rotation. However, half of the users mentioned that they are more likely to log their shoes into the system if they can receive personalized purchase recommendation.

“Purchasing shoes either based on my current collection or based on trends would be more valuable to me if I had to log that many shoes.”
  • Need for more direct item-logging experience and navigation

    Another user pointed out that the process of logging her shoes could be shorter. Some users also mentioned that they were confused about the gesture-based navigation throughout the usability testing.

  • See more value in recommendations for clothes rather than shoes

    More importantly, a user talked about that she would consider using the product only if it could give clothing recommendations as well. This reaction supports our initial idea of the digital closet product. With three weeks left before pitching the product, we decided to continue with the footwear direction for more validations on our hypothesis before we tackle clothing.

“This product might be more helpful if it’s for my clothes because I can’t see all of them, but less for my shoes because I can see them all.”
Final Results

Tailored recommendations

  • One recommendation at a time
    To minimize the cognitive load for daily footwear decision, we feature only a single shoe suggestion when the user opens the app in the morning.
  • Offer flexibility to adjust accordingly
    Users can swipe the card left or right to see more recommendations of the day. They can also swipe up to see all the different considerations (colors, weather, occasions, etc.) for this option and adjust according to their situation.
  • Improving with user feedback
    Simple would also ask the users for positive or feedback for the previous recommendation to improve the prediction accuracy.

Personal Digital Shoe Rack

  • Digestible collection
    To keep the footwear collection organized and visually consistent, we populate the digital shoe rack with shoes from a database. Also, the shoe rack is designed in only black and white to showcase the colors of the shoe collection.
  • Distilled shoe logging process
    Users add a pair of shoes from the database with existing shoe details (colors, material, function, etc.). They can also add notes to mark the condition of the shoes.
  • Shoe stats
    With the user feedback system, users can browse the statistics of particular shoes to check if they have overworn them.

Actionable discovery for shoe trends

  • Keeping up with the trend via a personalized newsfeed
    Based on the user’s collection and preference, the discovery feature will show the appropriate fashion trends and new release. Meanwhile, part of the revenue will come from brands sponsored posts in the discovery feed.
  • Purchasing from the discovery
    Users can buy the shoes that they like straight from the articles or reviews as well. The new purchase will be automatically added to the shoe rack to create a seamless experience. Another part of the revenue will be a transaction fee for each purchase.
Final Feedback

Like the piloting solution but struggles to relate to the problem

Pitching to the class with the addition of judges from the industry, we got some mixed feedback. While they appreciated the design execution, some judges were concerned about whether picking a pair of shoes in the morning is a problem worth solving. However, they acknowledged that it could be a solid pilot to prototype before moving into clothing in the future. Also, Simple was voted for one of the People’s choice awards within the class.

Simple was voted for one of the People’s choice awards within the class.
Retrospective

Minimum viable product vs. minimum “lovable” product

Scoping down to shoe rack help us illustrate the value of our products with more details within the given time. However, to make a minimum lovable product, Simple has to expand its capability to an entire closet. For many people, picking a pair of shoes to wear are not a problem because it is normally the last step to finish their outfit combination. The majority of the cognitive load is located during the other outfit decision. Therefore, after demonstrated the basic functions, this product will have to continue developing to a minimum lovable product to realize its market potential.

Group photo with the TAs and Ken Sandy, course instructor and VP of Product Management at MasterClass