Privacy-Compliant Tracking for a Health & Wellness Brand
Health & Wellness D2C · May - June 2026
A D2C health brand sellingsupplements, ayurvedic formulations, therapeutic products ran its growth on Meta and Google.
In early May 2026, Meta restricted all pixel data collection from the domain. The platform classified the product catalogue as health-sensitive and suspended all tracking saying that they were violating its data privacy policies. Purchase events, Page Views, and pixel-based audiences were all cut off overnight.
Blended ROAS collapsed from ~2.8x to ~1.1x.
Meta's domain classification is policy-driven and persistent. There was no way to simply turn the pixel back on.
The Problem
Since all pixel events were restricted.
→ There was no conversion signal in Ads Manager.
→ Algorithm had nothing to learn from.
→ Ad spend continued
→ Audiences decayed
→ Attribution went dark.
The brand needed a structural tracking reset, that would work in compliance with Meta's health data policies while keeping performance marketing intact.
BooleanMaths Solution
We implemented a Five-part privacy-compliant server-side tracking and decisioning architecture
Server-Side Event Tagging
All tracking moved off the browser pixel and onto BooleanMaths' CAPI infrastructure. Events were processed server-side before forwarding, giving the brand full control over what data reached Meta's servers.
URL Scrubbing
All URLs were stripped of product-level identifiers - category names, slugs, SKU codes before transmission. The referrer chain sent to Meta contained zero health-intent signals.
Generic Event Naming
Standard purchase and funnel events were replaced with privacy-safe, intent-neutral names. No Purchase event was transmitted. Generic signals kept the pixel active without leaking health context.
Hashed Customer Data Only
All customer PII shared with Meta was hashed. No product or purchase details were associated with the records. Meta received only anonymised identifiers for audience matching.
Attribution reset
Since purchase events no longer appeared in Meta Ads Manager, BooleanMaths data tracked server-side and stitched across browsers became the sole attribution source. All Ad campaign decisions moved to the BooleanMaths platform.
Results - Three Phases leading to Recovery
Phase 1 - Up to 7 May 2026
Normal Operations - Blended ROAS ~2.8x
Meta pixel was firing normally. Purchase events flowing. Audiences updated in real time. Performance fully measurable within Meta Ads Manager.
Phase 2 - 8 to 25 May 2026
Meta Domain Block - Blended ROAS collapses to ~1.1x
Meta restricts all pixel events. No conversion signal in Ads Manager. Algorithm flies blind. Ad spend continues with zero feedback loop. Audiences degrade. Attribution goes dark.
Phase 3a - 26 May to 15 Jun 2026
BooleanMaths Tracking Live - Blended ROAS recovers to 2.0–2.4x
Privacy-compliant CAPI goes live. Generic events resume data flow to Meta. BooleanMaths pixel becomes the sole attribution source. All performance reporting shifts to the BooleanMaths platform.
Phase 3b - 16 to 22 Jun 2026
BM Campaigns Scaling - Full measurement across Meta and Google
Brands start running individual campaigns using BooleanMaths audiences for exclusions and lookalikes, and making scaling decisions using True ROAS, and metrics on BooleanMaths dashboards while running the existing setup in parallel for comparison.
Campaign Performance - BM vs Standard
All data is calculated using BooleanMaths linear attribution.
The Hidden Loss BooleanMaths Made Visible
The account's largest campaign - Rs. 2,36,168 spend over the period - was running at 0.60x true ROAS. Without BooleanMaths' independent attribution, this was completely invisible: Meta's platform had no conversion data to report. BooleanMaths surfaced a Rs. 2.4L per month bleed and enabled an immediate pause decision that would otherwise have been impossible.
Current State
Privacy-compliant tracking is working smoothly, enabling the Brand to launch and test new creatives without restrictions.
Meta data flow is restored and the pixel went from fully blocked to live, while passing privacy-compliant signals.
BooleanMaths pixel replaced Meta's native reporting as the authoritative attribution layer across Meta and Google.
BM campaigns outperformed standard Meta campaigns by 2.9x on True ROAS - 2.00x vs 0.70x and CPA Rs. 546 vs Rs. 1,587.
New customer acquisition maintained at 69%, confirming privacy constraints did not compromise audience quality.
2.00x True ROAS on BM Campaigns
3.43x Best Campaign ROAS
₹2.4L Monthly Loss Surfaced
70% New Customers on BM Campaigns
Takeaway for Health Category Brands
Meta's health data restrictions are not a temporary edge case. They are a structural constraint for any brand in health, wellness, pharmaceutical, or personal care and they will expand as privacy regulation tightens globally.
BooleanMaths gave these brands three things Meta's native tooling can not:
Privacy-compliant pipeline that kept data flowing under restrictions;
Independent attribution layer that kept performance legible when Meta's dashboard went dark;
Audience infrastructure that let BM campaigns significantly outperform the broader account.
Server-side, privacy-compliant, independently attributed - this is the baseline model for performance marketing in sensitive categories.