Analyzing yShade user interactions with
an AI model

My roles

UX Researcher

Meet the team

[ +1 ]

Lead UX Designer

Project timeline

May 2024 - January 2025

Looking For a thorough breakdown?

Email for a private report of this study

yShade's users utilize AI to find beauty products.

How do the users feel about this?

yShade.ai is a beauty tech company that has introduced its users, who are in search of tailored beauty products, to the AI model: Aiysha.

For this study, we aimed to understand users' interactions with Aiysha on a mobile platform. Through this application they are able to discover and purchase beauty products tailored to them by Aiysha's inclusive dataset. Usability testing was conducted with five participants using a mid-fidelity prototype of the yShade application as the test subject.

Research question

How do our users navigate through the discovery and purchasing phases of their conversations with Aiysha?

[ 1 ]

Design Study

[ 2 ]

Test Application

[ 3 ]

Analyze Results

[ 4 ]

Identify Next Steps

Their response?

Users were intimidated… but excited!

"

[Aiysha] is a good start to the makeup-finding process.

"

-Interview Participant

"

Where does contour fit?

"

-Interview Participant

Discovering products…

Participant Ratings:

6.8 / 9

Lacked Guidance

Users did not feel confident interacting with AI. This raised the concern of providing more info transparency and guidance for product comparison.

Confusing to navigate

Users struggled to understand their personalized list of beauty products. The info hierarchy of this screen made users feel lost when looking at the options.

Helpful Recommendations

The process of receiving product suggestions based on the users' provided information was viewed positively. Users liked the idea of conversing with premade text prompts in their chats with AI.

"

[It's] very simple and straightforward.

"

-Interview Participant

Purchasing products…

Participant Ratings:

8.2 / 9

'Add to Cart' Doubts

Adding items to the cart was difficult for users due to a misunderstanding of the button's label.

Understood Results

Iconography for the cart was recognizable by the users, creating an easy experience of looking at their selected products to purchase.

How does this help to improve Aiysha?

The data collected from this study leaves us with two guiding questions:

➡ How might we build users' confidence in AI?

➡ How might we improve the product discovery feature's navigation?

By identifying these pain points, cross-team collaboration is streamlined to address users' concerns and increase confidence in the use of AI models in the beauty tech industry.

© 2025 Kathryn Theisen