Virtual Fitting Room
Role: UX Researcher, UX Designer
Stakeholders: Research Team, Design Team, Product Team
Duration: 3 weeks
Deliverables: Research artifacts, stakeholder presentations
Background: This case study covers a two-part project spanning a user study of a novel 3D fitting room tool and a subsequent redesign. For the duration of this project, I was mentored by members of both the UX research team and the product design team.
The problem
Size and fit are significant barriers for users purchasing clothing online. Sizing notations across brands are often inconsistent and incongruent, lowering customer confidence in their appropriate size.
Without accurate and accessible information on size and fit, users are disincentivized from purchasing and abandoning the item. To address this issue, URBN (parent company of Urban Outfitters, Anthropology, and Free People) wanted to explore new fit tool options. This project consists of a study exploring style.me, a virtual fitting room tool that allows users to generate a 3D avatar, and a subsequent redesign addressing style.me’s usability errors.
The goal of this project is two-fold: to understand what style.me does well for customers regarding fit and how customers interact with novel fit tools and to conceptualize an optimized version for URBN's platforms.
The people
Raliee is a plus-size woman who is an avid online shopper. She often finds it difficult to find her size and usually buys multiple sizes, returning the extras in-store.
Liv is an older woman based in the UK who is looking for new summer dresses. She doesn’t often buy online, but Anthropologie’s new dress line caught her eye.
Raliee and Liv are both personas drawn from past research at URBN, representing the customer journeys of both returning and new shoppers. They are the users in this study and for whom the style.me redesigns will be tailored.
The Research
To guide the study, I met with key stakeholders from the product team to determine the highest priority concerns regarding style.me. With this information, I consulted with my collaborators to create 4 main research questions.
Does using the style.me avatar try-on feature give users more information than what already exists on a PDP regarding size & fit?
Are there any usability issues or points of friction experienced while using this tool, specifically with building an avatar, “trying on” different sizes, etc?
How well does this tool solve the user’s need to determine what size to buy, and how that size will lay on their body (fit)?
Does using the style.me avatar try-on feature give users more information than what already exists on a PDP regarding size & fit?
Research Questions
Usability Task & User flows
For the task itself, users were recruited from UserTesting and given a product landing page (PDP) of a retailer who had already integrated style.me. They were then prompted to explore the page and identify their size in whatever way they wanted. If they did not click into the widget, they were then prompted to do so. From there they followed two main flows.
Since the goal of the study was to gather insights into how users interact with size and fit information, this study was designed to be qualitative with the intention of thematic analysis. To gather data, I selected seven guiding questions for the user interviews.
What questions, if any, do you have about the way this dress would fit on you, e.g. the way it would lay on your body?
How confident are you that the size you would purchase would be for you?
How easy or difficult was it for you to use this tool? What was easy or difficult about it?
Did you adjust the body shape, face, and hair options? Why or why not?
Did this tool give you any additional information about what size you would purchase?
Did this tool give you any additional information on how that size would fit your body?
How confident are you that the size you would purchase would work for you? Are you more or less confident than you were before you used the tool?
Evaluative Questions & Metrics
My co-researcher and I independently reviewed the interview transcripts and developed a coding schema of a high inter-rater reliability. We then each coded the interviews, synthesizing key insights in response to the evaluative questions and overall usability.
Widget Usability
Technical Issues: A portion of users were prevented from using the tool at all due to technical issues; these users were excluded from feedback about the usability of the tool. Overall 42% experienced serious lag and 18% were not able to complete tasks.
Discoverability: Style.me had low discoverability. Users' initial impressions showed CTAs were hard to notice and deprioritized compared to the size guide and TrueFit CTAs (42% vs 100% call-out rate). 55% of users were unable to find the secondary flow which would allow users to see different sizes on the avatar even after being prompted to do so.
Accessibility: Users found the prompts intuitive and had no issues will filling out their measurements, however, 38% of plus size users hit a weight limit with the sliders despite the shop offering larger sizes that would fit per the size guide/TrueFit.
Size and Fit
Improved confidence in size choice: Half of the participants expressed that the Style.me tool validated, but did not determine, their sizing choice. Rather, users looked to reviews and past experience to make an initial size selection.
Avatar did not address fit concerns: Due to uncertainty with the accuracy of the 3D avatar, most users still had specific concerns about the fit of the item (i.e they would be able to wear it, but did not know if the arms would be a little too tight or if it was too short).
High cognitive load: While many users found style.me provided additional validation about size, many expressed that the tool was simply not their first choice due to being long and complicated.
The graphic below contextualizes UX of style.me in comparison to the existing size guide tool.
Findings
The ReDesign
User research of style.me identified several ways in which a virtual fitting tool could be used by customers to increase confidence in their size choices; however, there were several significant issues that prevented users from effectively engaging with this tool to improve experiences with online size and fit. In particular, there were major issues with the widget’s discoverability on the PDP and of the secondary user flow displaying fit information. As part of this project, I was tasked with re-designing the tool to better address these usability concerns.
Design goals
Design Analysis
To identify specific design choices that may have contributed to high-level usability issues, I revisited user recordings and observations to understand what aspects caused friction and what improvements can be made to address them.
In this redesign, there are two gateways into the virtual fitting room designed to help with the issue of low discoverability. The primary’s CTA is right next to the size guide and uses the terminology "find my fit" to help users better identify the end goal of what they are engaging with (as opposed to previous CTAs "3D Try on"). The secondary CTA is located at the end of the image carousel and more clearly identifies the option to customize your avatar and see the item on you.
Accessing Style.me widget
To address users' inability to find their recommended size, they will not only get a pop-up that they must click elsewhere to dismiss but also have a persistent designation of it (via asterisk). To allow users to better access fit information, I pulled in the "fit details" (bust, waist, hips) from the original secondary flow into the primary flow. In addition, users will be able to navigate between sizes without having to click into another screen, as was previously designed.
Displaying Size and Fit
Final Screens
Reflections
This project presented an incredible opportunity to not only learn the ropes of conducting a user study but also to effectively transform that research into informed design decisions. Wearing the hats of both researcher and designer allowed me to observe the dynamics between various teams and identify distinct communication requirements for my work. Additionally, it was very rewarding to contribute to an outward-facing, customer-centric website and engage directly with end users.