The Problem: Running Blind on Three Channels
A US-based direct-to-consumer brand selling premium outdoor gear was running paid ads on:
- Google Ads (Search + Shopping)
- Meta Ads (Facebook + Instagram)
- TikTok Ads
- LinkedIn Ads (B2B gifting segment)
Google Analytics and their attribution model showed a blended ROAS of 1.8x. The Google and Meta campaigns had tracking. TikTok and LinkedIn did not.
Their marketing team believed TikTok and LinkedIn were "awareness channels" with no direct return. They were about to cut both.
Before making that decision, they brought in Brand Growth Hack for a full tracking audit.
What the Audit Revealed
The Google Analytics data was incomplete. The attribution model was Google Analytics 4's default "Data-Driven Attribution" — but DDA can only attribute conversions it knows about, and it only knows about sessions where the GA4 tag fires.
For TikTok and LinkedIn, there was no pixel. GA4 wasn't attributing any revenue to these channels. Every sale that originated from a TikTok or LinkedIn click that later converted was being attributed to "direct" or "organic" — depending on the last touch.
When we audited the customer journey data through Shopify, we saw something significant: 34% of customers had multiple sessions before purchase. And the UTM analysis on those multi-session customers showed large gaps — sessions with no UTM, no referrer, no identifiable source.
The hypothesis: TikTok and LinkedIn clicks were initiating purchase journeys that were completing via direct navigation or Google search. Standard last-click attribution was giving Google and Meta full credit for sales that TikTok and LinkedIn had started.
The Solution: Cross-Platform Pixel Deployment
Step 1: TikTok Pixel via GTM (Days 1–4)
Deployed TikTok Pixel through Google Tag Manager:
- Base code via Custom HTML tag, firing on All Pages
- Standard events:
ViewContent,AddToCart,InitiateCheckout,Purchase - Each event includes
content_id,value,currencyfrom the Shopify DataLayer - Purchase events include order ID for deduplication
Additionally configured TikTok Events API (server-side) via Stape.io for iOS attribution recovery — same approach as Meta CAPI, but for TikTok's Events API.
Step 2: LinkedIn Insight Tag + Conversion Events (Days 5–8)
LinkedIn Insight Tag deployed via GTM:
- Base tag on All Pages
- Conversion event: Purchase (with order value)
- Added LinkedIn-specific audience segmentation: tracks which job titles and company sizes were purchasing (invaluable for the B2B gifting segment)
Step 3: Pinterest Tag (Days 9–10)
The brand was running organic Pinterest content but no paid ads. We deployed the Pinterest tag speculatively — to collect data ahead of any future paid activation. Within 3 months, they activated Pinterest Ads using the audience data already collected.
Step 4: Reddit Pixel (Days 11–12)
Same approach as Pinterest — deployed speculatively for future audience building in their outdoor gear community subreddits.
Step 5: Unified DataLayer (Days 13–18)
All four new tags were connected to a unified DataLayer event structure:
- Purchase events push: order value, product category, product IDs, customer email (hashed)
- All pixels receive identical data — no platform gets inconsistent numbers
- Centralised in GTM, so any data layer change propagates to all platforms simultaneously
Step 6: Cross-Platform Attribution Model (Days 19–25)
Built a custom Looker Studio attribution dashboard connecting:
- Shopify order data (source of truth)
- Google Ads conversion data
- Meta Ads conversion data
- TikTok Ads reporting API
- LinkedIn Ads reporting API
The dashboard used first-touch attribution (which channel initiated the customer journey) alongside last-click attribution (which channel received credit in GA4) to identify cross-channel attribution gaps.
What the Data Showed After 45 Days
With full tracking across all channels, the picture changed completely:
| Channel | GA4 Revenue (Before) | True Revenue (After) | Difference |
|---|---|---|---|
| Google Ads | $180,000 | $175,000 | -$5,000 |
| Meta Ads | $145,000 | $148,000 | +$3,000 |
| TikTok Ads | $0 | $78,000 | +$78,000 |
| LinkedIn Ads | $0 | $42,000 | +$42,000 |
| Total | $325,000 | $443,000 | +$118,000 |
The "$120K in hidden revenue" was split between TikTok ($78K) and LinkedIn ($42K) — two channels the team was about to cut.
The true blended ROAS across all channels: 5.0x (not the 1.8x GA4 was reporting).
Google and Meta had been receiving credit for sales that TikTok and LinkedIn had initiated. The "underperforming" channels were actually the ones driving new customer acquisition. Google and Meta were capturing the bottom-of-funnel conversions from those audiences.
Outcome
The brand:
- Increased TikTok Ads budget by 3x after seeing true ROAS of 4.8x
- Increased LinkedIn Ads budget for B2B gifting targeting by 2x
- Activated Pinterest Ads using the audience data collected speculatively
- Adjusted their attribution model to first-touch for new customer acquisition decisions
Within 6 months of the full tracking deployment, monthly revenue had grown from $325K to $510K — attributed partly to better budget allocation guided by accurate cross-platform data.
Key Insight
Multi-channel attribution is not a "nice to have" for brands spending across multiple platforms. Without cross-platform pixels, you are making budget allocation decisions based on incomplete data — and inevitably, you will underfund the channels that are actually driving growth.
The universal tracking setup cost less than one month of wasted TikTok and LinkedIn budget. The attribution clarity it provided was worth far more.
