The Challenge
A leading KSA e-commerce brand was facing a classic problem: high traffic but low conversion rates. Their website was showing the same products to every visitor.
Note: This case study is illustrative of the results AI personalization delivers for KSA e-commerce brands.
The Solution: AI-Powered Personalization Engine
Working with Conneqt, the brand implemented a three-layer personalization strategy:
Layer 1: Homepage Personalization Each visitor's homepage is now dynamically assembled based on previous purchase history, browse behavior, demographic signals, and real-time inventory data.
Layer 2: Category Page Optimization Product rankings within categories are personalized based on individual purchase history.
Layer 3: WhatsApp Re-engagement Customers who viewed a product 3+ times without purchasing receive a personalized WhatsApp message with product details, reviews, and a time-limited offer.
The Results (6 Months Post-Implementation)
| Metric | Before | After | Change |
|---|---|---|---|
| Conversion Rate | 1.2% | 3.1% | +158% |
| Average Order Value | 890 SAR | 1,340 SAR | +51% |
| Revenue (monthly) | Baseline | 3x Baseline | +200% |
| Cart Abandonment Rate | 78% | 61% | -22% |
| WhatsApp Recovery Rate | N/A | 34% | — |
Key Learnings
- Arabic-first AI matters. Models trained on Arabic purchase data outperformed generic models by 40%
- WhatsApp is the recovery channel. Email recovery had 8% open rates; WhatsApp had 94%
- Speed matters. Real-time personalization (under 100ms) outperforms batch-computed recommendations by 2.3x