Wasim Ullah

• 60% — CPA Reduction
• 2x — ROI Improvement
• 4 — GCC Markets Managed Simultaneously

AXA Gulf’s digital campaigns were generating leads at high cost with inconsistent performance across the UAE, Bahrain, Kuwait, and Oman. There was no unified forecasting framework, campaigns lacked negative keyword management, and creative fatigue was degrading results, making it impossible to allocate budgets efficiently across markets.

Rebuilt the campaign architecture market by market, introducing audience segmentation layers, negative keyword suppression, automated bidding models, and quarterly performance forecasting. Multi-market Looker Studio dashboards enabled real-time budget reallocation. ASO was added to reduce paid install dependency and diversify acquisition channels.

Within 6 months, CPA dropped 60%, and ROI doubled across all four markets. Forecasting models were adopted by senior leadership for annual planning, elevating marketing’s strategic influence within the organization.

• $1+ Million — In Attributable Revenue
• 40% — Conversion Rate (vs. 8–15% Industry Average)
• 10x — Performance Improvement via A/B Testing
• 60% — Reduction in Cost Per Acquisition

BuyAnyInsurance needed to scale lead acquisition in a competitive UAE insurance market while improving conversion rates and reducing reliance on expensive paid acquisition. Campaigns targeted a bilingual (English/Arabic) audience across car, health, and travel insurance, with no reliable analytics infrastructure or conversion optimization framework in place.

Redesigned the conversion architecture from the ground up. Implemented GA4 and GTM event tracking to capture granular user behaviour, then used Microsoft Clarity heatmaps and 100+ iterative A/B tests across landing pages, CTAs, and form flows. Simultaneously deployed an AI outbound calling system using n8n, RetellAI, and GoHighLevel CRM to automate renewal outreach at scale. A bilingual GEO strategy was added to capture AI search visibility across ChatGPT, Gemini, and Perplexity.

CVR reached 40%, nearly 3x the 8–15% UAE insurance industry benchmark. The system delivered $1+ million in attributable revenue. A/B testing program achieved a 10x performance improvement. The AI outbound calling system handles 1,000+ contacts daily.

• 1,000+ — Contacts Processed Per Day at Peak
• 25% — Reduction in Manual Sales Effort

BuyAnyInsurance had thousands of lapsed insurance leads in GoHighLevel CRM being manually re-engaged — an unsustainable process that couldn’t scale without significant headcount increases. Relying on human resources would waste time and money.

Designed a 3-workflow n8n automation architecture: a daily cron workflow extracting lapsed leads from GHL and triggering RetellAI voice agent calls at controlled concurrency; a webhook handler processing call outcomes and updating GHL contact records with tags, notes, and custom fields; a weekly retry workflow for follow-up sequences. A tag state machine enabled intelligent segmentation and phased rollout from 10 to 1,000 daily contacts.

The system scaled from 10 to 1,000+ contacts per day while reducing manual sales effort by 25%. The tag-based segmentation (llr, ai-called, ai-renewal-captured/ai-llr-followup, ai-lost-lead) created a self-managing pipeline.

• 100% — Patient Scheduling Funnel Visibility Achieved

Eagan Counseling Clinic needed accurate tracking of the patient scheduling journey to understand drop-off points and improve online appointment bookings, a critical revenue driver for the practice, with no existing analytics infrastructure.

Set up GTM and integrated with GA4, architecting detailed event tracking across each step of the scheduling funnel. Configured custom GA4 exploration reports visualizing the full patient journey and identified key drop-off points. Reports were designed for the clinic’s non-technical marketing team to use independently.

The clinic gained full visibility into the patient scheduling funnel for the first time, identifying specific drop-off points for CRO prioritisation. The reporting setup was designed for non-technical users.