Inspire Watch — GA4 + Server-Side Tracking Rebuild | Case Study

Inspire Watch was scaling ad spend — but Shopify said one revenue number, GA4 said another, and Ads Manager claimed a third. Browser pixels were losing a significant share of conversions to iOS, Safari and ad blockers, so Meta and Google were optimizing on incomplete data. Growlyze rebuilt the entire tracking stack first-party: GA4 configured properly, tracking moved server-side, and every purchase delivered to the ad platforms with correct deduplication.

Catalog: 3,000+ products — smartwatches, fitness and health trackers, luxury, analog and digital
Platform: Shopify, shipped worldwide from multiple warehouses
Growth: Paid-led, via Meta and Google
Engagement: Analytics & Server-Side Tracking

Overview

Inspire Watch is a US-based watch brand on Shopify, selling men's, women's and kids' watches across smartwatches, fitness and health trackers, luxury, analog and digital styles — a catalog of 3,000+ products shipped worldwide from multiple warehouses.

Growth is driven by paid social and search, which makes accurate conversion tracking the difference between scaling and burning budget.

The GA4 Home report showing event count, active users, total users and purchases for the last 30 days, with a trend line comparing the period against the previous one. — The GA4 property after the rebuild — events, users and purchases reporting from one configured source.

Challenge

Four ways the data was lying.

Pixels were missing conversions. Browser-based tracking silently lost purchases to iOS, Safari and ad blockers — real orders the ad platforms never saw, and could never learn from.
GA4 was misconfigured. Missing item-level data, duplicate purchase events, and channels lumped together — the full journey from view_item to purchase could not be followed cleanly.
Three platforms, three numbers. Shopify, GA4 and Ads Manager each reported different revenue. Nobody knew which was right — so nobody could confidently make budget decisions.
Algorithms learning from fragments. Meta and Google built lookalikes and bidding from a fraction of real buyers — budget flowed to the wrong audiences, and every scaling decision was a coin flip.

Solution

Tracking rebuilt from the ground up — first-party.

GA4 done properly. The full e-commerce event chain — view_item, add_to_cart, begin_checkout, purchase — with item-level data, clean channel groupings, and duplicate events eliminated.
Server-side GTM. Tracking moved off the customer's browser into a server container the brand owns. First-party, resilient to iOS and ad blockers — with faster page loads as a side effect.
Meta CAPI and enhanced conversions. Purchase data sent server-to-server to Meta and Google with correct deduplication — so every real order reaches the ad algorithms.
Consent Mode v2. Cookie compliance and regional privacy handling configured correctly — accurate data without legal exposure, for a store selling worldwide.
One attribution report. A Looker Studio report reconciling Shopify orders, GA4 and the ad platforms — one version of the truth for the whole team.
Validated end to end. Test orders traced through every platform before handover, plus documentation so the team owns the system.

GA4 reports side by side: event count by event name led by page_view, user_engagement, view_item and session_start; key events by event name including purchase; and items purchased by item name. — The event chain reporting cleanly — page_view through view_item to purchase — with key events separated from raw traffic and item-level data attached.

Attribution by channel

With purchases arriving server-side and channel groupings cleaned up, value could finally be traced back to the channel that earned it — rather than pooled into an unattributed bucket that made every budget conversation a matter of opinion.

GA4 reports showing event count by event name, key events filtered to purchase, and average 120-day value broken out by first user primary channel group across unassigned, email, organic search, paid social, organic shopping and direct. — Average 120-day value split by the channel that first acquired the user — the report that turns budget debate into a decision.

Worldwide, measured

Inspire Watch ships worldwide, so the rebuild had to hold up across regions and privacy regimes at once — Consent Mode v2 configured correctly, with reporting that stays accurate wherever the customer is.

GA4 audience reports: active users by country shown on a world map with a ranked country list, alongside user activity over time comparing 30-day, 7-day and 1-day active users. — Active users by country and rolling activity — consistent reporting across every market the store sells into.

Impact

Same ads. Same budget. Better data — better results.

+27% more conversions attributed, with event match quality up to 8.6/10 from roughly 5.
CPA fell 18% within weeks, with zero creative changes — the algorithms simply learned from every real buyer instead of a fraction of them.
One report the whole team trusts. With reconciled attribution, budget moved on evidence rather than debate, and campaigns claiming credit for organic sales were exposed.