Mina Awadallah / Projects

Unlocking the Power of Product Discovery for online shoppers with AI

Exploring how AI can empower users, provide discovery based products results with the goal to reduce the reliance on faceted filters

Lead product designer

Experiment, MVP

SEO, Growth, Engineering

Mina Awadallah / Projects

Unlocking the Power of Product Discovery for online shoppers with AI

Exploring how AI can empower users, provide discovery based products results with the goal to reduce the reliance on faceted filters

Lead product designer

Experiment, MVP

SEO, Growth, Engineering

Mina Awadallah / Projects

Unlocking the Power of Product Discovery for online shoppers with AI

Exploring how AI can empower users, provide discovery based products results with the goal to reduce the reliance on faceted filters

Lead product designer

Experiment, MVP

SEO, Growth, Engineering

Intro

Background

I led an innovative design hackathon, focusing on the integration of AI into our product design process within the design team. We delved into understanding our customers' brief interactions with our platform, aligning our efforts with the company's strategic objective to develop a mobile application. During this initiative, I explored potential AI applications in eCommerce & search to enhance user experience and streamline product discovery.

Intro

Goals

Enhance usability

Discover new ways to redesign the filter component for mobile, that is more contextual to user search and focus more on relevant product filters.


Personalized experience

Encourage users to create accounts, through direct messages , dedicate space to discover relevant products,

Encourage exploration

Encourage users to discover trends in fashion seasons, following the company's major drive to focus on seasons.

Enhance usability

Discover new ways to redesign the filter component for mobile, that is more contextual to user search and focus more on relevant product filters.


Personalized experience

Encourage users to create accounts, through direct messages , dedicate space to discover relevant products,

Encourage exploration

Encourage users to discover trends in fashion seasons, following the company's major drive to focus on seasons.

Problem Statement

Ladenzeile has identified usability challenges in our current filter system, which is crucial for efficient product exploration and discovery, and as a product comparison platform, These challenges have become more apparent, By addressing these issues, My aim was to redefine our filter functionality, making it more intuitive and user-friendly to improve satisfaction.

Discovery

Features Definition

Based on Ladenzeile's earlier user research data, I aimed to tackle challenging components in terms of usability such as filters, User Explorations in terms of curated personalized content and relevant products, and the company goal of retention

To Improve user journey by providing a shorter path to narrow down their search results and provide a personalized user experience towards the goal of user retention.

Hypothesis

User adoption vs Current journey

User preferences to search by manual input would be shifted to a more efficient way using Voice or natural language while initiating search as they would do in a physical store.

Optimizing interaction

With a focus on exploration and discovery filters are collapsed by default, however dynamic contextual filters based on their initial input would be revealed on interaction.

Style by look (Glamour View)

Allowing users to upload images of their favorite outfits either wearing them or a photo of similar outfits, would help the user discover more and easily compare styles.

User adoption vs Current journey

User preferences to search by manual input would be shifted to a more efficient way using Voice or natural language while initiating search as they would do in a physical store.

Optimizing interaction

With a focus on exploration and discovery filters are collapsed by default, however dynamic contextual filters based on their initial input would be revealed on interaction.

Style by look (Glamour View)

Allowing users to upload images of their favorite outfits either wearing them or a photo of similar outfits, would help the user discover more and easily compare styles.

Design Process

Key Activities

I presented the prototype to key decision-making stakeholders such as the head of Frontend, Director of Product, and Engineering Director), from there I asked for a small task force to be formulated voluntarily and two engineers signup, the following key activities took place.

Feedback Loops

In the early stage of the project, I went through multiple feedback loops with different teams to better understand the feasibility of the solution such as Supply and demand considering they are the key stakeholders as well as introducing the concept of a white label project to utilize our 15 million product inventory.


Identifying current database structure

I pitched the project as a solution that wouldn’t require alteration to the current database on the contrary it would allow optimization due to the usage of AI, we proved the concept of using the current database as it and the AI-generated accurate results.

HTML Prototype following Figma Prototype

Presenting an early Figma prototype steered enough curiosity for the stakeholders, so I developed a basic prototype showing the functionality of one of the touch points to get a buy-in, the engineers took over at a later stage to implement open ai API and test out the search functionality and voice input.

Feedback Loops

In the early stage of the project, I went through multiple feedback loops with different teams to better understand the feasibility of the solution such as Supply and demand considering they are the key stakeholders as well as introducing the concept of a white label project to utilize our 15 million product inventory.


Identifying current database structure

I pitched the project as a solution that wouldn’t require alteration to the current database on the contrary it would allow optimization due to the usage of AI, we proved the concept of using the current database as it and the AI-generated accurate results.

HTML Prototype following Figma Prototype

Presenting an early Figma prototype steered enough curiosity for the stakeholders, so I developed a basic prototype showing the functionality of one of the touch points to get a buy-in, the engineers took over at a later stage to implement open ai API and test out the search functionality and voice input.

Feedback Loops

In the early stage of the project, I went through multiple feedback loops with different teams to better understand the feasibility of the solution such as Supply and demand considering they are the key stakeholders as well as introducing the concept of a white label project to utilize our 15 million product inventory.


Identifying current database structure

I pitched the project as a solution that wouldn’t require alteration to the current database on the contrary it would allow optimization due to the usage of AI, we proved the concept of using the current database as it and the AI-generated accurate results.

HTML Prototype following Figma Prototype

Presenting an early Figma prototype steered enough curiosity for the stakeholders, so I developed a basic prototype showing the functionality of one of the touch points to get a buy-in, the engineers took over at a later stage to implement open ai API and test out the search functionality and voice input.

Design Process

Prototype & Results

The design process went through multiple iterations and a feedback cycle and the following design represents the outcome of those activities.

User experience

Internal surveys feedback that it felt natural, while some have concerns over voice and image input, the redesigned filter components were perceived as user-friendly and accessible and were set to be scheduled for usability testing and design surveys.

Search results

After providing the AI with our database set in json format and providing a manual input “ I’m looking for summer clothes or i’m looking for a night party dress for this weekend “ the AI managed to identify our database with its current structure and provide results.

User experience

Internal surveys feedback that it felt natural, while some have concerns over voice and image input, the redesigned filter components were perceived as user-friendly and accessible and were set to be scheduled for usability testing and design surveys.

Search results

After providing the AI with our database set in json format and providing a manual input “ I’m looking for summer clothes or i’m looking for a night party dress for this weekend “ the AI managed to identify our database with its current structure and provide results.

Let’s talk

Email @ mina@awadallah.io

Hire me over Upwork

Reach out @