⚠️ Illustrative example — all figures and names here are fictional. It exists only to show the required case-study format.
E-commerce · السعودية · Jan — Mar 2025 (12 weeks)
Increase direct sales for a perfume store while lowering cost per order without cutting spend.
- Meta Ads
- Google Shopping
- GA4
01
Business Challenge
The store spent 70% of budget on cold-audience “sales” campaigns, so cost per order was high and ROAS unstable. The real problem was not the ad: the product page revealed shipping cost late, and 40% of users dropped at cart.
02
Strategy
The funnel was restructured into three layers: • Existing demand: Google Shopping for high-intent keywords. • Discovery: Meta with creative explaining fragrance concentration differences (the audience did not know them). • Recovery: Dynamic Product Ads for cart abandoners. A 20% weekly creative-testing budget was ring-fenced before any scaling.
03
Execution
Full tracking was rebuilt on GA4 + Meta Pixel, and the purchase event was fixed (it fired twice). Campaigns were then restructured: one campaign per funnel layer, one ad set per creative concept, with six angles (price, longevity, occasion, gifting, comparison, usage proof).
04
Optimization
Week 3: paused four ad sets running 60% above the target CPA. Week 5: shifted the largest budget to the “gifting” angle after it produced the highest purchase completion. Week 7: expanded lookalikes in 1% → 3% steps while monitoring frequency. Week 9: switched off the recovery campaign because its return did not cover its cost.
05
Results
| Metric | Value | Definition |
|---|---|---|
| Cost per order | SAR 38.4 | Total spend ÷ confirmed orders over the same 12-week window. |
| Conversion rate | 2.9% | Confirmed orders ÷ campaign-sourced sessions only (not all site traffic). |
| ROAS | —pending | Attributed campaign revenue ÷ campaign spend, 7-day click attribution window. |
عرض البيانات كجدول
| أسبوع 1 | 71 |
| أسبوع 2 | 64 |
| أسبوع 3 | 58 |
| أسبوع 4 | 52 |
| أسبوع 5 | 47 |
| أسبوع 6 | 44 |
| أسبوع 7 | 43 |
| أسبوع 8 | 41 |
| أسبوع 9 | 40 |
| أسبوع 10 | 39 |
| أسبوع 11 | 39 |
| أسبوع 12 | 38 |
Attribution note: Total store revenue exceeds attributed campaign revenue, because part of sales came from returning direct customers — so store-wide ROAS is not claimed here.
06
Key Learnings
The biggest gain was not inside the ads manager: showing shipping cost earlier on the cart page lifted completion more than any targeting change. Lesson: audit the funnel before blaming the campaign.