The Automation Failure
To drive ad listings and lead generation at a massive scale, Bayut & Dubizzle required full-funnel personalization. However, deep-dive technical audits revealed a systemic failure: their underlying data triggers were fundamentally flawed across 227 distinct events. They had invested heavily in Machine Learning (ML) engines, but because the event data feeding those engines was broken, their personalization efforts were failing to convert.
Repairing the ML Data Pipeline
- The Enterprise Trigger Repair: We methodically mapped and repaired 126 user attributes and fixed broken deep links across the entire platform, restoring the integrity of their data collection.
- The ML Integration Architecture: We built the exact Jinja logic and infrastructure required to feed their internal ML engines directly into MoEngage Catalogs, finally unlocking the ability to serve highly accurate property and car recommendations at scale.
Double-Digit CVR Growth
The infrastructure repair delivered immediate, massive ROI. Bayut & Dubizzle achieved a 35% conversion rate on their Welcome Journeys and a 29% conversion rate on Upsell Journeys (successfully nudging free users to highly profitable VAS products). Push Notification conversions skyrocketed to between 23% and 50%.
Is bad event data breaking your automated workflows and ML models? Request an e-CENS Data Audit to find and fix the broken triggers in your funnel.
