GearUp

GearUp is an e-commerce app featuring a voice shopping assistant and Generative AI-powered prototyping and personalization to increase conversion rates.

Role
UI/UX Designer

Employer
Passion Project

Platform
iOS

Duration
15 Weeks

Background

I set out to address how to help users efficiently research and make quicker decisions when purchasing a high-ticket item, such as an e-bike.

As the sole Product Designer on the project, I conducted end-to-end research, usability testing, prototyping, and multiple iterations to develop the GearUp app flow.

Initial Problem Discovery

User Interviews

I interviewed five young professionals aged 25-40 to understand their shopping behaviors for high-price purchases.
The findings revealed key factors influencing decisions on high-ticket items, including customizable bicycles, and highlighted differences between in-store and online buying experiences.

Approach

How might we reduce pre-cart abandonment and increase conversion rates?

I focused on the following KPIs:

• Task Completion Rate: Measure the percentage of users who successfully add items to their cart without frustration.
• Conversion Rate: Increase the percentage of users who add items to their cart, signaling a higher likelihood of repeat usage and ongoing engagement with the app.

Ideate

Through research, I found that there is a growing preference in users leveraging Voice Assistant and Generative AI technologies. These features were aimed at streamlining the shopping experience and minimizing user friction.

71%

of users prefer voice search, driven by younger consumers, households with children, and high-income households. (Source PWC)

75% 

of consumers expect personalized experiences and are increasingly open to using AI and voice-assisted tech to streamline their shopping journey. (Source Salesforce)

10x

improvements in conversion rate when personalization is delivered more sophisticatedly to users. (Source Adobe & Incisiv)

High Fidelity Prototype

I conducted three rounds of testing and iterations with five users, ultimately landing on this final high-fidelity prototype. Key updates include addition of a chat-type option, a reduction in the steps required to view final recommendations, enabling users to more effectively compare and search for products, and offering greater flexibility for users to customize their orders after receiving the initial suggestions.

Impact

+25% in Conversion Rate with
User Satisfaction increase toward the shortened research process and time, easy to complete tasks, and ability to customize further.

Users are inclined to continue to use this app to help find products and make purchases. 

Challenges

Developing a new Natural Language Model (NLM) and collecting customer data for full personalization is a significant challenge.

Some users suggested exploring plug-ins to streamline this process.