How we rescued the client’s analytics and gave back 30% of its revenue visibility

Context

The client is a Portuguese e-commerce store selling gym and fitness equipment online across Europe. The brand had been running paid media campaigns across Google Ads and Meta Ads, and relying on GA4 as its primary analytics and attribution layer.

The work was delivered by Better Data Today on a white-label basis for Aero Agency, the client’s marketing partner. All client-facing communication was handled through Aero Agency.

Challenges

When Aero Agency brought us in, the client’s GA4 data was significantly undercounting both transactions and revenue. The store’s back-end order data and GA4 were telling different stories, and ad platforms were optimising based on the wrong one.

What the data showed (April 2026 – May 21^st 2026):

Note: these percentages are calculated against the client’s back-end records.

Missing 34% of revenue in the analytics layer meant:

• Google and Meta were optimising campaigns with corrupted conversion signals.
• Audience segments and lookalike models were built on incomplete purchase data.
• ROAS reporting in ad platforms was artificially inflated, so true performance was hidden.
• The agency had no reliable way to attribute revenue to campaigns or channels.
• Consent mode modelling was producing estimates instead of working with clean first-party data.

Root causes

After auditing the Shopify and Google Tag Manager (GTM) setup, we identified the following issues:

• GA4 purchase event: it was firing inconsistently due to Shopify checkout flow conflicts and order confirmation page timing issues.
• Meta Pixel: browser-only implementation with no server-side redundancy, meaning any ad blocker or ITP restriction silently dropped the event.
• Consent mode: it wasn’t properly configured, so consent denials were dropping events entirely instead of routing through modelled conversions.
• Server-side GTM: it was not in place, leaving all tracking fully exposed to browser-side loss.
• dataLayer: purchase events were not reliably populated with order value and product data on all order confirmation paths.

Our solution

We delivered a full tracking rescue across four service areas (GA4, server-side GTM, Meta CAPI, and Consent Mode v2), designed to make the client’s data resilient, consent-compliant, and reliable for paid media optimisation.

1) GA4 implementation

We rebuilt the GA4 event tracking from the ground up on Shopify, as follows:

• Audited and corrected the purchase event trigger to fire reliably on all order confirmation paths, including post-purchase upsell pages.
• Validated the dataLayer push to ensure order_id, value, currency, and item-level data were consistently populated.
• Configured GA4 Enhanced Ecommerce to capture the full funnel: view_item, add_to_cart, begin_checkout, purchase.
• Set up conversion events in GA4 and linked them to Google Ads for campaign optimisation.

2) Server-side GTM

We implemented a server-side GTM container to remove browser-side dependencies for critical conversion events:

• Deployed an sGTM container on a first-party subdomain, removing reliance on third-party script delivery.
• Routed GA4 and Meta conversion events through the server-side container, making them independent of ad blockers and browser restrictions.
• Implemented Cookie Keeper / first-party cookie persistence to maintain attribution across ITP-restricted browsers (Safari).

Server-side events now provide a stable fallback regardless of client-side script failures.

3) Meta Conversions API (CAPI)

We added a server-side Meta CAPI implementation to complement and deduplicate the browser Pixel:

• Events now fire via both browser Pixel and CAPI simultaneously, with event deduplication handled through matching event IDs.
• Advanced Matching parameters (email, phone) are passed via CAPI to improve audience matching quality.
• Purchase, InitiateCheckout, and AddToCart events are all covered by CAPI.
• Signal quality score validated in Meta Events Manager after go-live.

4) Consent Mode v2 & GDPR compliance

We configured a fully compliant Consent Mode v2 setup compatible with Shopify and the Portuguese regulatory environment:

• Integrated with the store’s CMP to pass consent signals correctly to GTM, GA4, and Google Ads.
• Configured ad_storage, analytics_storage, and ad_user_data signals per user consent state.
• Enabled GA4 and Google Ads modelling to recover estimated conversions for users who declined consent, maintaining campaign optimisation continuity.

Consent-denied events no longer drop silently, since they are now modelled.

Results

We validated results by comparing the client’s actual store order data against GA4 recorded data, day by day, across the full period specified in the table below. The improvement was immediate and measurable.

1) Tracking quality (before vs. after)

* Revenue capture rate: the remaining ~4% gap is within the expected tolerance for consent mode modelling and browser-side ad blocking. No further fix is required.

2) Agency impact (for Aero Agency)

From the agency’s perspective, the tracking rescue delivered immediate operational and commercial value:

• Client reporting is now based on reliable data.
• Campaign optimisation signals improved across both Google Ads (via GA4 conversions) and Meta (via CAPI), giving the paid media team accurate data to work from.
• The consent mode implementation protects the client’s compliance posture under Portuguese and EU GDPR requirements.
• Server-side infrastructure is now in place and reusable for future tracking needs without rebuilding from scratch.