MARCHEMIX
A 3.2x ROAS after switching to server-side tracking
Talk about my projectFlow: Shopify store, to sGTM (Stape), then Google Ads, Meta CAPI.
3.2x
ROAS over the engagement
Source: Google Ads and Meta Ads, over the 11-month engagement
11-month mission on a Shopify store — full GTM tagging overhaul sGTM deployment via Stape Meta CAPI with deduplication then full acquisition management. Result: 3.2x ROAS over the mission duration.
3.2x ROAS · sGTM Stape + Meta CAPI deployed · Attribution made reliable
- Context
- E-commerce · Paris · Shopify · Growth partner
- Services
- Tools
01 — The context
The existing tracking was broken — wrong attribution misconfigured events barely usable GA4 data. Impossible to properly optimize campaigns without reliable data. The tracking overhaul was the prerequisite for any performance.
02 — Before / after
| Before | After |
|---|---|
| Broken GTM tagging · misconfigured events | Full GTM tagging overhaul (dataLayer · GA4 · Google Ads · Meta) |
| Uncertain attribution on Google Ads and Meta side | sGTM deployed via Stape — tags migrated server-side |
| No server-side tracking · degraded signal post-iOS14 | Meta CAPI implemented with event deduplication |
| Campaigns optimized on wrong data | Search Shopping PMax campaigns managed and optimized |
03 — The result
3.2x ROAS over 11 months — direct result of reliable attribution combined with active campaign management. Server-side tracking restored the Meta signal post-iOS14 and improved lookalike audience quality.
04 — My role
Growth partner over 11 months — rebuilt tracking, then managed acquisition on an ongoing basis (Search, Shopping, PMax).
05 — What I took away
Moving to server-side tracking (sGTM/Stape) has become a prerequisite, not an option, to restore Meta signal quality post-iOS14.
