★★★★★ Real client work · Audited and fixed

Deep-dive case studies from real tracking rebuilds.

These are not generic examples. Each case study comes from actual client audits, delivery reports, and technical fixes across Elevar, GTM, Shopify, GA4, Meta CAPI, Klaviyo, and Google Ads.

Shopthuli overhaul Monday Muse recovery Mandset migration blueprint

The core problem

The store had a premium tracking setup, but legacy apps were suffocating it. Native Shopify Google/Facebook apps, Triple Whale, Analyzify, manual pixels, and Elevar were all trying to own the same events. Instead of one clean source of truth, every platform was receiving conflicting signals.

What the audit found

  • Shopify Customer Events had multiple pixels active at once, creating duplication risk across Meta and TikTok.
  • GTM was holding stale ecommerce values, creating a phantom $212.50 GA4 purchase on a $0 exchange.
  • Revenue was mapped to subtotal instead of grand total, creating a $250 reporting gap.
  • GA4 was receiving data from three different Google tracking IDs, fragmenting reporting.
  • Meta Purchase was not registering correctly and CAPI was active in name only — critical events were still browser-only.
  • Klaviyo had duplicate flow-trigger risk from Elevar and native Shopify integrations firing together.

Before

  • Triple tracking conflict across Shopify, GTM, and native apps.
  • GA4 revenue inflated and product-level stats broken.
  • Meta had dual pixels, poor EMQ, and no reliable server-side payloads.

After

  • Redundant tracking apps removed and Elevar restored as the single source of truth.
  • GTM revenue mapping rebuilt and tag sequencing corrected.
  • Meta CAPI restored with deduplication and server payload processing.

What was fixed

Shopify
Removed redundant Google/Facebook app tracking and Customer Events conflicts.
GTM
Fixed ghost revenue, subtotal mapping, session-reset race conditions, and server/browser conflicts.
GA4
Consolidated tracking into a single property and restored product-level item tracking.
Meta
Restored standard Purchase, improved EMQ, removed dual-pixel counting, and forced server-side payloads through Elevar.
Klaviyo
Stabilized the server-side webhook path and flagged duplicate native flows for client-side draft action.

The core problem

Meta was receiving inconsistent signals for the same user actions. Some events had mismatched values, some contents arrays included zero-value free gifts, and some landing-page server signals arrived too late to match the browser event.

Issue 1 — Purchase value mismatch

  • The browser pixel sent VAT-inclusive totals including shipping.
  • Elevar CAPI correctly sent ex-VAT product subtotal.
  • Meta saw two values for the same order ID and flagged deduplication errors.
  • The fix aligned the browser pixel value to match Elevar’s server-side CAPI value.

Issue 2 — BOGOS free-gift contamination

  • Free gifts like “The Bag” and “Pill Box” entered event contents arrays with a value of 0.00.
  • Meta rejected those contents arrays because item_price/quantity/id requirements were violated.
  • The BOGOS flow also caused Add to Cart to fire up to three times for one action.
  • The fix filtered zero-value products before they reached Meta and blocked duplicate Add to Cart firing.

Issue 3 — ViewContent timing and enrichment

  • The server tag fired at 4.5 seconds while the browser event fired at 2.0 seconds.
  • Fast paid traffic left before the server signal fired, creating browser-only gaps.
  • The fix rewired the server tag to fire the moment the dataLayer push completed.
  • Hashed email, phone, first name, last name, fbc, IP address, and User Agent were added where available.

Before

  • 67% of purchases flagged for value mismatch.
  • Zero-value BOGOS gifts contaminated Meta contents arrays.
  • ViewContent had timing gaps and weak enrichment.

After

  • Browser and server purchase values aligned.
  • Free gifts stripped from funnel contents arrays where technically possible.
  • Browser and server ViewContent signals fired together with stronger matching data.

Technical nuance

Not every issue was blindly “fixed” with a patch. For the Shopify purchase webhook, the report clearly flagged the platform limitation: Shopify sends every line item server-side, which means some free-gift filtering cannot happen at the webhook layer. That distinction matters because it shows the fix is scoped around what the platform can actually control.

The migration risk

Changing tracking directly can damage live bidding algorithms if conversion actions are duplicated, broken, or switched too quickly. The migration needed to protect Simprosys data while Elevar was validated.

Technical plan

  • Audit existing GTM and Simprosys setup before touching live tags.
  • Verify Elevar product IDs match Google Merchant Center feed IDs to protect Shopping campaign attribution.
  • Pass full item arrays into Google Ads purchase events for product-level reporting.
  • Use Enhanced Conversions to improve modeling for suppressed search-term and attribution data.

Safe rollout sequence

  • Create Elevar Google Ads conversions as a new separate conversion action.
  • Set Elevar to Secondary / observation only while Simprosys stays Primary.
  • Run both side-by-side for several days without letting Google’s algorithm optimize against Elevar yet.
  • Once Elevar numbers are verified clean, switch Elevar to Primary and move Simprosys to Secondary.

Validation requirements

  • Elevar Server Monitor screenshots.
  • GTM Preview Mode confirmation.
  • GA4 DebugView testing.
  • Google Ads Diagnostics confirming Enhanced Conversions.
  • Consent Mode V2 sync against the existing Swedish/EU cookie banner.

Bad migration

  • Flip everything at once and risk duplicate conversions.
  • Break bidding history before new data is proven.
  • Assume Consent Mode is working without testing it.

Safe migration

  • Audit first, stage Elevar secondary, compare numbers, then switch primary.
  • Protect bidding data while validating server-side accuracy.
  • Confirm product IDs, item arrays, Enhanced Conversions, and Consent Mode.

Want this level of diagnosis on your store?

Send the store URL and I’ll review the tracking symptoms before recommending the right scope.

Get free audit →