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)

MetricBeforeAfterChange
Conversion Rate1.2%3.1%+158%
Average Order Value890 SAR1,340 SAR+51%
Revenue (monthly)Baseline3x Baseline+200%
Cart Abandonment Rate78%61%-22%
WhatsApp Recovery RateN/A34%—

Key Learnings

  1. Arabic-first AI matters. Models trained on Arabic purchase data outperformed generic models by 40%
  2. WhatsApp is the recovery channel. Email recovery had 8% open rates; WhatsApp had 94%
  3. Speed matters. Real-time personalization (under 100ms) outperforms batch-computed recommendations by 2.3x