AI-Powered Web App for Multi-vendor E-commerce Solution
Anamoul Rouf
Web Designer
Product Designer
UI & UX Design
Figma
Miro
Notion
Product: Web Application for Multi-vendor E-commerce Solution
Timeline: March 2023 – August 2023.
My Role: UX Research, UI/UX Design
Tools: Miro, Adobe XD, Zeplin, Figma, Notion.
Platform: Web Application.
Team Collaboration: Product, Marketing, Support, and Tech.
Background
The multi-vendor e-commerce landscape is evolving, and AI is the game-changer here. This platform aims to simplify and enhance the buying and selling process. Sellers often face hurdles in setting up their stores, managing listings, and reaching the right customers. Buyers, on the other hand, want personalized, efficient shopping experiences.
By integrating AI, this platform solves both ends: from simplifying product uploads with auto-generated descriptions and price suggestions to offering buyers tailored search results and recommendations.
Goals
Our main challenge was to design a web application that would empower sellers, delight buyers, enhance buyer-seller communication, and leverage insights.
Empower Sellers: Provide tools for seamless product uploads, tag generation, and price optimization using AI.
Delight Buyers: Ensure personalized shopping experiences with smart product recommendations and a robust search system.
Unify Buyer & Seller Roles: Create a hybrid dashboard that allows users to switch between buying and selling functionalities seamlessly.
Enhance Communication: Enable efficient buyer-seller messaging and a notification system for updates and insights.
Leverage Insights: Deliver an advanced reporting system to give users data-driven insights into their activities and performance.
Success Metrics
Increase in the number of active sellers by simplifying product upload processes.
Improved buyer engagement through personalized recommendations and faster search results.
Higher conversion rates due to optimized pricing and detailed product descriptions.
The adoption rate of hybrid dashboards by multi-role users (buyers/sellers).
Positive user feedback on the messaging and notification features.
Design Challenge
Hybrid Dashboard: Balancing the dual roles of buyers and sellers in a single, intuitive interface.
AI Implementation: Making AI features like product tag generation and price suggestions seamless and user-friendly.
Insights and Reporting: Designing a system that’s both visually appealing and provides actionable insights for users.
Communication: Ensuring messaging feels natural and efficient without overwhelming users.
Scalability: Future-proofing the platform to handle growth in users and features.
The UX Process: The Lean UX process to move forward
When we have minimal time and resources, selecting an efficient, sustainable, and flexible design process is always challenging. I have chosen the approach to the Lean UX process. It is an iterative process with three simple steps.
Think
Make
Check
...and keep repeating the steps.
Lean UX Process
The Discovery: The discovery phase was a quick, high‐intensity effort
The discovery phase was a quick, high‐intensity effort that allowed us to -
Interest in AI:
Users see value in AI for tasks like automating mundane processes (e.g., generating product descriptions or price recommendations) and for improving efficiency.
The curiosity stems from AI’s ability to save time and potentially increase profitability.
Target Users:
Small to medium-sized businesses, individual sellers, and buyers who also dabble in selling.
These users often manage dual roles, requiring a platform that caters seamlessly to both buying and selling activities.
Non-Intimidating Features:
The success of these tools depends on creating a user-friendly interface where AI feels like a helpful assistant rather than a daunting system.
Adoption Hinges on Simplicity:
Despite the interest, there’s hesitation or reluctance if the AI tools feel too complex or technical.
Many users, especially individual sellers or SMBs, may not have advanced tech expertise. Overcomplicating AI features could alienate this audience.
The Findings: We Visited Our Targeted Users
We have sessions with 20 targeted users. We also had brainstorming discussion sessions with our internal cross-functional team members. So what we found-
Simplified Product Management for Sellers:
Sellers often struggle with manually creating product descriptions, tags, and pricing. The platform's AI aims to alleviate this by automating these tasks, saving time and reducing errors.
This simplification not only increases efficiency but also encourages more sellers to actively use the platform.
Personalized Experiences for Buyers:
Buyers prefer platforms that provide tailored recommendations and anticipate their needs.
AI's ability to analyze preferences and past behaviors helps reduce decision fatigue, leading to faster and more satisfying purchases.
Enhanced Communication:
Communication between buyers and sellers is often treated as secondary but is crucial for building trust and resolving issues.
By incorporating a robust messaging feature, the platform ensures smoother transactions and fosters user satisfaction.
Effective Notifications:
Notifications need to strike a balance between being timely and relevant without overwhelming the user. The thoughtful design of this system keeps users engaged and informed without causing frustration.
These findings provide a foundation for addressing user needs while leveraging AI to deliver a seamless and intuitive experience for both buyers and sellers.
Research: The Competitive Analysis
AliExpress:
Popular for its global reach and affordability, but its interface can be overwhelming for new users, and it doesn't cater well to dual-role functionality or personalized recommendations.
Shopify:
Excellent tools for sellers but lacks robust AI product suggestions. The platform is easy to use for storefront creation but doesn't address the buyer-seller hybrid needs directly.
Amazon Marketplace:
Strong recommendation engine with a vast marketplace, but its complexity and lack of flexibility for individual sellers to switch between buyer and seller roles make it less appealing for small businesses.
Etsy:
Great for niche markets with unique products, but it offers limited advanced reporting and lacks strong AI-powered tools for search or product management.
Idation & Testing: Sketches, Wireframing, and Testing
The following steps were to sketch wireframes based on the user flow maps and build the flow and low-fidelity prototype for a series of testing sessions.
Creating wireframes helped me to visualize the ideas and gave us the basic structure for how users would use the app.
When creating these wireframes, we kept the user needs from the earlier findings and research in mind.
I have done several testing sessions with a close group of users.
UI Design: Design Guideline & Design System
Initially, we built basic style guidelines. Afterward, we built a highly customizable design system to create a design language.
UI Design: User Interface Design
Before starting the UI design process, I always built a highly customizable design style guideline and reusable components.
I have designed a design style guideline and reusable components.
Exprience the UI Design Screens
Testing & Validation: Usability Testing
The goal was to assess the app's overall usability and identify areas for improvements that could facilitate the completion of the tasks. The participants were given the same set of jobs.
Due to the NDA issue, I've only shared the information authorized by the authority.
Refine & Deploy: We refined and updated the design & deployed it for development
Based on the usability testing sessions, we refined and updated the design. Then, we created deliverables for development handover. I maintained continuous communication with the developers during the development period.
After the successful deployment, it is time to collect feedback from the user.
Results & Future Iteration: It is a One-time Project
As it was a one-time client project, I didn’t have the opportunity to measure the results.
Flow for an ideal design process.
However, I always prefer to follow a mixed approach of Design Thinking and the Lean UX design process for continuous product development.
Thank you :)
