AI-Powered UX Research:
Informing the Redesign of Casablanca Market’s E-Commerce Experience
Project overview
Casablanca Market, an e-commerce brand specializing in Moroccan cuisine and home décor, transitioned from a physical retail store to a fully online marketplace. The goal was to create a seamless, engaging shopping experience for U.S. customers while elevating small Mediterranean sustainable brands, ultimately positioning Casablanca Market as the go-to platform for Mediterranean artisanal products.
Despite increasing online traffic, the brand faced three critical challenges:
High bounce rates – Many users left within the first 15 seconds.
Low conversion rates – Checkout abandonment remained a major issue.
Engagement issues – Users spent little time interacting with the site.
The challenge was to translate the immersive in-store experience into a compelling digital one while uncovering the barriers preventing users from engaging and converting. A clear need emerged for research to better understand user needs and pain points, ensuring alignment with business goals.
My Role
Founding Product Designer: Strategy, UX Research, Usability Testing.
Collaborators
Lean team: CEO, Marketing Manager, Lead Software Engineer, Product Manager.
Tools
Miro, Chat GPT, Gemini, Perplexity, Poll the People, User Persona
Defining KPIs
The primary business goals over the next six months were to:
Increase conversions by 20%
Reduce bounce rates by 15%
Achieve at least 50% usability satisfaction scores
Deliver visible improvements (2-5%) within one month
Outcome
I designed an AI-enhanced UX research strategy that integrated AI-driven insights with traditional UX methods. This hybrid approach enabled rapid identification of usability issues, quick design iterations, and early MVP improvements. The results:
Conversions
42% boost in conversion rates from improved navigation and checkout flows.
Bounce rates & engagement
35% reduction in bounce rates, leading to better user engagement.
User satisfaction
Significant improvements in mobile usability, increasing customer satisfaction by 35%. 4.8 average usability test ratings for new designs.
Visible improvements
52% increase in sales within 2 months of MVP design implementation.
AI-Driven UX Research: A Hybrid Approach
To uncover why users disengaged and abandoned purchases, I developed a holistic research strategy combining AI-powered research, traditional UXR methods, and usability testing. This allowed for faster, deeper, and more actionable insights.
AI-Powered Problem Space Exploration
I leveraged AI tools to synthesize industry reports, market trends, and sentiment analysis from competitor reviews, significantly reducing research time. Competitor benchmarking provided insight into common UX pitfalls and best practices in e-commerce design.
Using AI for problem space exploration and competitive auditing, I found that AI was particularly useful in synthesizing existing online information about overall industry trends and UX challenges. However, when it came to competitor analysis, AI failed to identify our key competitors. Still, it proved valuable by uncovering additional indirect competitors worth considering. AI also helped create a structured benchmark competitor UX audit plan and checklist, enabling us to conduct more targeted and scoped research moving forward.
Next, to further validate gathered information, I I conducted a User Behavior Analysis (Hotjar) using heatmaps and user session recordings.
This research confirmed that Casablanca Market struggled with common e-commerce pitfalls and revealed:
-
Upselling opportunities – The "frequently bought together" feature performed well and could be expanded.
-
Navigation Issues might be responsible for low engagement and drop offs: e.g. non-sticky menu bar—users struggled to access the menu, search, and cart.
-
Disruptive pop-ups—users often exited when the email collector appeared.
-
Mobile responsiveness problems – UI inconsistencies frustrated mobile shoppers.
AI-Enhanced Moderated User Research & Usability Testing
Using AI-generated personas, I simulated user research interviews and usability studies before real-world testing. This process helped:
-
Identify blind spots in the research and refine interview questions.
-
Optimize usability tests, eliminating bias and redundancies.
-
Predict user behaviors and preemptively address bottlenecks before conducting live tests, saving time and costs.
Key findings:
Usability testing confirmed and expanded upon previous research, revealing:
-
Mobile responsiveness issues were a major conversion blocker.
-
The pop-up email collector disrupted the experience and led to early exits.
-
CTAs and navigation elements needed stronger visual emphasis—users overlooked key actions.
-
Homepage confusion: Many users initially mistook Casablanca Market for a recipe blog.
-
Limited search discoverability—users struggled to find products and categories.
-
Inconsistent product pages—missing care instructions, unclear "Sold Out" indicators, and illogical "frequently bought together" pairings.
Additional UX Gaps Identified:
-
Navigation & Search Issues – Users had difficulty finding products, recipes, and returning to the homepage.
-
Product Page Inconsistencies – Care instructions and reviews weren’t consistently displayed.
-
Checkout Issues – "Sold Out" items were still being added to the cart, disrupting the checkout flow.
-
Accessibility Concerns – Some blurry images and the homepage slideshow negatively impacted UX, particularly for users with astigmatism and ADHD.
-
Positive Feedback: Users appreciated sustainability values, artisan support, and fair-trade messaging, though the information could be better structured.
-
Some users also appreciated relevant product suggestions, unboxing videos, or recipe recommendations on food-related pages, but remarked lack of consistency.
Research Synthesis & Stakeholder Presentation
I leveraged AI tools to transcribe user interviews, enabling quick and accurate documentation of insights. Using AI-driven pattern recognition and clustering techniques, I identified recurring themes, pain points, and user needs, which streamlined the discovery of key UX issues. AI-assisted sentiment analysis helped prioritize user concerns, allowing for a more data-driven approach to problem-solving.
For synthesis, I employed AI tools to categorize and visualize user feedback, creating affinity diagrams that mapped pain points, user behaviors, and potential opportunities. This helped structure the research findings into digestible insights, making it easier to pinpoint high-impact areas for improvement.
During the brainstorming phase, I used AI-powered ideation tools to generate solution concepts, evaluate feasibility, and compare potential design directions. By leveraging AI-driven mind mapping and impact-effort matrices, I was able to prioritize UX improvements based on urgency, feasibility, and user impact.
To facilitate stakeholder alignment, I used AI to design, conduct a structured workshop. We also recorded our session and used AI tools to generated summaries and insights.
Based on the workshop discussions, we developed the following implementation plan:
P0 (Critical Fixes)
Fix mobile responsiveness issues immediately.
Improve the checkout experience by ensuring "Sold Out" items can’t be added to the cart.
Implement sticky navigation for menu and search.
Enhance search functionality—make products, collections, and recipes easily discoverable.
P1 (High-Priority UX Enhancements)
Redesign the homepage to clarify the value proposition and declutter the layout.
Improve CTAs, Labels, Improve accessibility—ensure buttons and labels (e.g., "Sold Out") are prominent & compliant with accessibility standards.
Ensure branding and design consistency across the platform by implementing a design system.
P2 (Medium-Priority Enhancements)
Refine the email pop-up experience to reduce disruption.
Enhance product and collection pages to ensure consistency in descriptions and care instructions.
Optimize upselling features—expand the "frequently bought together" model.
Redesign the "Add to Cart" flow for a smoother experience.
Create a structured Recipe Page template to integrate recipes seamlessly into the shopping experience.
Design a merchant onboarding flow to streamline partnerships with new artisans.
Implementation Plan
-
Prioritize mobile UX improvements first—this was the most significant usability concern.
-
Keep the web platform structure intact while adding necessary pages.
-
Make minimal, high-impact design changes first to maximize efficiency within time and budget constraints.
-
Gradually redesign user experiences to improve retention, increase user satisfaction, and reinforce Casablanca Market’s brand presence as a top-tier e-commerce platform.
Key Takeaways
This research revealed that Casablanca Market’s biggest UX barriers were navigation complexity, mobile responsiveness issues, unclear product offerings, and disruptions in some user flows (add to cart & checkout). Addressing these concerns immediately instantly boosted engagement and conversions.
Key Takeaways:
AI accelerates UX research but requires human oversight. AI-powered persona simulations and usability test predictions helped optimize research focus, but human analysis was essential to empathise with the users, interpret results and prioritize actions effectively.
UX improvements don’t have to be drastic to be impactful. Simple changes—such as making CTAs clearer, improving search functionality, and reducing homepage clutter—can immediately enhance user engagement and conversions.
Mobile-first design is non-negotiable. With an increasing percentage of users on mobile, responsiveness issues directly impacted revenue. Fixing these should remain a top priority.
By combining AI-powered research with traditional UX methodologies, we accelerated discovery, implemented quick wins, and laid the groundwork for a scalable, user-friendly e-commerce experience.