Total Traceability: From Instagram to Purchase (Without Data Loss)

Migrated to a clean GTM architecture to fix critical attribution errors between Instagram and payment gateways. We filtered data pollution and enabled granular tracking, turning blind guessing into actionable insights in just 2 weeks.

Laptop screen displaying a Google Tag Manager workspace with a 'Success: Tags Fired' overlay, confirming active GA4 and social tags. A bamboo plant sits in the background on a wooden desk, representing Bambú Studio's clean data approach.

Analytics (GA4 + GTM Logic) — 2 weeks — Data Integrity 100%

Context (The blind spot)

Carabachuis was generating sales but lacked visibility on how. Their measurement relied on Google Site Kit, injecting heavy, redundant code that slowed down the store.

Worse, the data was polluted: internal team visits (staging/admin) mixed with real customer data, and the payment gateway (Mercado Pago) was hijacking attribution. This broke the tracking of Instagram campaigns, making it impossible to calculate realcalculate  ROAS.

What we did — Orchestration & Cleaning

  • Goodbye Site Kit, Hello GTM: We migrated the entire measurement stack to Google Tag Manager. We centralized scripts, cleaned the source code, and improved load times by removing bloated plugins.
  • True Attribution (Referral Exclusion): We configured GA4 to ignore mercadopago.com and mpago.la as traffic sources. Now, if a user arrives via an IG Story, the sale is correctly attributed to Instagram, not the bank.
  • Data Hygiene: We filtered out internal traffic and staging environments. The reports now reflect 100% real customer behavior.

What we did — Micro-landing Strategy (/hola)

Carabachuis uses a custom "Linktree-style" page (/hola). The challenge was measuring clicks on specific buttons (WhatsApp, Packs, Games) without stripping the UTM parameters the user brought from Instagram or printed QRs.
  • Logic Persistence: We implemented custom logic in GTM to capture and persist entry_source and entry_medium parameters throughout the session.
  • Granular Events: We created custom events (cta_click, contact_whatsapp) to determine which button converts best, distinguishing whether the user came from a physical QR code or a digital link.

Data & Measurement (Key configs)

  • Standardized UTMs: Defined a strict taxonomy for campaigns (source/medium/campaign).
  • Enhanced E-commerce: Granular tracking of view_item, add_to_cart, begin_checkout, and purchase with values in ARS.
  • Cross-domain tracking: Ensured session continuity across the purchase journey.
Split-screen showing technical optimization results: a Mobile PageSpeed Insights score of 92/100 alongside a Google Analytics 4 admin panel configuration excluding 'mercadopago.com' from referral traffic to fix attribution errors.

Results (From guessing to knowing)

  • Funnel Visibility: Crystal clear GA4 reports showing exactly where users drop off (Home → Product → Cart → Payment).
  • Web Performance: Significant reduction in head blocking scripts (faster load times).
  • Data-Driven Decisions: The client now knows exactly which Ad campaign delivers real ROI and which ones simply drive window shoppers.

What’s next (Optimization)

  • Visualizing the Impact: Deploying a Looker Studio dashboard to give stakeholders a real-time, simplified view of KPIs without needing to log into GA4.
  • Audience Activation: Building specific segments (e.g., "Cart Abandoners" or "High-Value Users") to feed Google Ads remarketing campaigns.
  • Continuous Improvement (CRO): Using the new accurate data to run A/B tests on the checkout flow and improve conversion rates month over month.
 

Interesting articles

Ready to stop guessing where your sales come from?

Bad data kills growth. We fix attribution errors and clean your reports so you can scale with confidence.