Product Discovery & Search Optimisation

The Problem

Traditional keyword search caused high cart abandonment from customers unable to find relevant products. Static rule-based recommendations couldn't personalise to individual behaviour. Business stakeholders had no agreed KPI framework to measure search improvement, making it impossible to validate any change.

My Contribution

I defined the search and discovery requirements by facilitating workshops with product managers, merchandising teams, and data analysts. I documented current-state search failure modes through data analysis and user research, established a standardized KPI framework across the business, and translated behavioral data patterns into recommendation engine requirements. I designed the A/B testing criteria and acceptance thresholds used to validate each improvement before full rollout, and authored the measurement plan that tracked conversion impact post-launch.

The Solution

Requirements-led discovery overhaul: stakeholder workshops to define failure modes, KPI taxonomy standardization, technical requirements for search and recommendation engines, A/B test design, and a structured measurement plan for post-launch validation.

Results

• 35% lift in product discovery conversion
• 18% increase in average order value
• 28% lower search results bounce rate
• 40% faster time-to-find (3+ min → <2 min)
• 5M+ search queries daily at sub-100ms latency

Tech Stack