Server-side tracking · GA4 · Meta CAPI · sGTM & Stape
Recover lost ad attribution, fix Meta CAPI deduplication, and build server-side GTM pipelines that don’t break when browsers, iOS, or ad-blockers change. For e-commerce and lead-gen brands that need their numbers to match reality.
Typical first step: a fixed-scope tracking audit of one property — what’s firing, what’s dropping, and the exact fix list.
The symptoms of broken measurement — and what they’re costing you on every ad dollar.
Browser-side pixels drop 20–40% of conversions. Your ad platforms optimize on bad data and your ROAS reads low.
Browser pixel and server event both fire with no shared event_id — Meta counts the sale twice and inflates ROAS ~2x.
No hashed user data in the CAPI payload, so match quality floors and the algorithm can’t find buyers.
GA4, the ad platforms, and Stripe all disagree, and nobody on the team trusts the weekly report.
A server container that owns your tags — resilient to iOS and ad-blockers, faster pages, and first-party data you control.
Browser + server events sharing one event_id, hashed user data for high Event Match Quality, and no double-counting.
Clean event schema, key events/conversions, server-side purchase integrity, and Measurement Protocol where it belongs.
Server-side purchase events from Stripe webhooks into GA4, the ad platforms, and your CRM — logged and reconcilable.
Clean fields, list hygiene, and lifecycle data wired so marketing and sales see the same truth.
Google Enhanced Conversions and Consent Mode v2 configured so you stay measured and compliant.
A sample of real tracking and data-engineering builds. Full portfolio and reviews on Upwork.
sGTM on Stape + Meta CAPI with event-id dedup and hashed user data — recovered conversions browsers were dropping and restored trustworthy ROAS.
Apps Script handler turning Stripe webhooks into server-side purchase events, with a test suite and a sheet system-of-record behind it.
Rebuilt the CAPI payload with proper user_data hashing and dedup, lifting Event Match Quality and cutting phantom conversions.
Rebuilt a messy contact database into engagement × geography segments with suppression lists that protect sender reputation.
One property, fixed scope: I map what’s firing, what’s dropping, and the exact fixes — a document you can act on even if you build it yourself. Prefer platform escrow? Hire me on Upwork.