3,686% ROI from AI Fragrance Finder and Personalised Recommendation Engine
The Challenge
Fragrance Samples UK is a UK-based ecommerce retailer selling sample-size fragrances. They approached EXPRE to improve their conversion rate and automate their product recommendation process. With thousands of fragrances available, customers frequently suffered from choice overload, leading to a 72% cart abandonment rate - well above the industry average. The challenge was to reduce decision fatigue, guide customers to the right product confidently, and prove commercial value rapidly with a lean AI MVP.
Goals & Objectives
Increase conversion rate across all traffic channels. Reduce cart abandonment through AI-guided product discovery. Implement AI-powered personalisation to match customers to the right fragrances. Improve average order value through intelligent cross-sell recommendations. Automate email sequences to re-engage abandoned carts.
Our Approach
EXPRE built a custom AI Fragrance Finder tool using GPT-4o that guides visitors through a scent preference quiz and matches them to products from the full catalogue. The tool was integrated with WooCommerce for real-time product recommendations, stock-awareness, and dynamic pricing display. EXPRE also built automated email sequences for cart abandonment using n8n, triggered by user behaviour and personalised using the quiz data captured during browsing. The entire build was accelerated using Webgentic OS, EXPRE's AI-assisted development workflow, reducing build time from months to weeks.
Key Metrics
3,686% ROI on initial investment. 42% increase in conversion rate. 28% uplift in average order value. 60% reduction in bounce rate on product pages. Cart abandonment rate reduced from 72% to under 40%.
The Results
Live at fragrancesamples.co.uk. 3,686% ROI achieved within 6 months of launch. The AI Fragrance Finder now handles the majority of first-time visitor journeys, reducing bounce rate by 60% and lifting conversion across Google Ads, SEO, direct, and email traffic. Average order value increased by 28% as personalised cross-sells prompted basket additions. The platform continues to run continuous AI optimisation, improving recommendation accuracy with every purchase signal.