TikTok shows 1.9x ROAS.
Meta shows 3.8x.
One of them is folklore.
For founders who built the brand on TikTok and creators. Upload your free Google Analytics report. See which channels actually make you money, in minutes.
No signup for the demo · Full refund guarantee · GA4 CSV upload, 5–10 min insights · See full pricing · Book a 20-min call
The €120K question: is TikTok building your brand or just going viral?
You went viral. Orders came. Now you're spending €150K/year on TikTok, creators and Spark Ads, and finance wants receipts.
TikTok says 1.9x ROAS. Meta says 3.8x. But your whole funnel starts on TikTok. What if Meta is just collecting the sale TikTok already made?
The math:
TikTok "attributed" revenue: €150K × 1.9x = €285K
TikTok incremental revenue: test market shows a 31% drop = €120K lost
Difference: €120K of revenue TikTok drives but never gets credit for.
Cut TikTok = save €150K, lose €120K in revenue, and the For You page forgets you. Keep TikTok = explain "low ROAS" to finance forever.
You're stuck. Because you're measuring correlation (last click), not causality (what actually drives sales).
Twinkels found 28% of ad spend going to the wrong channels. See where yours should go.
Correlation vs causality: the difference is €120K
Traditional attribution (correlation):
Meta had the last click → Meta gets 100% credit → TikTok looks "low ROAS" → cut TikTok → the funnel dries up.
Behavioral intelligence (causality):
TikTok makes them want it → they check you out for a week → Meta retargeting closes → one 18-day journey → TikTok drives €120K incremental → keep it, scale it.
Incremental Sales = (Revenue with Channel) - (Revenue without Channel)
Traditional attribution: Measures correlation (which channel touched the customer last)
Behavioral intelligence: Measures causality (which channel drives incremental sales)
Confidence-scored results with data health indicators. Named Dutch TikTok-first brands already see their real numbers.
How Gen Z customers actually buy: the paths your dashboard hides
Path #1: TikTok → search → Meta
Sees the product on their For You page (Day 1) → Googles the brand (Day 6) → clicks a Meta retargeting ad (Day 12) → buys.
Last-click: Meta gets 100% credit. TikTok shows "low ROAS."
Causal analysis: In this example TikTok drives 35% of the sale. Cut TikTok = this customer never shows up.
Path #2: Creator video → TikTok Shop browse → email
A creator posts the honest review (Day 1) → they browse your TikTok Shop but don't buy (Day 3) → your welcome email closes it (Day 9).
Last-click: Email gets 100% credit. The creator shows zero ROI.
Causal analysis: In this example the creator drives 40% of the sale. Cut creators = the pipeline empties in weeks.
Path #3: TikTok → group chat → Instagram → Meta
The clip gets shared to a group chat (Day 1) → they check your Instagram (Day 4) → Meta retargeting closes (Day 10) → buys.
Last-click: Meta gets 100% credit. Again.
Causal analysis: TikTok started the journey and the dashboard logs none of it. That's the whole problem.
Pattern: Discovery channels (TikTok, creators, organic) show "low ROAS" in dashboards because they start journeys instead of ending them. In these worked examples they drive 30-40% of incremental sales. Cut them = the funnel dries up.
From upload to causal read
Setup time
Per causal read, no subscription
Why TikTok-first attribution is harder than it looks
The sale never happens where the journey starts
Gen Z discovers on TikTok, checks receipts on Instagram and Google, and buys through a retargeting ad or an email. Last-click hands all the credit to the closer and calls the opener worthless.
Creators do not come with a pixel
There is no tracking pixel on word of mouth. A creator video feeds Meta, Google and organic for weeks; the dashboard logs it as zero. Causal analysis measures the lift instead of the click.
Virality is not revenue
A two-million-view video feels like winning. Views are not money. The only question that matters: did revenue move because of it? That is a causal question, and dashboards cannot answer it.
Finance speaks a different language
You know the brand runs on TikTok. Finance sees a 1.9x ROAS line and wants it cut. A causal read turns 'trust me' into a number.
Bottom line: TikTok-first brands live at the start of the funnel, exactly where last-click sees nothing. Traditional attribution fails hardest here. Causal analysis works best here.
Brands already see their real numbers
The Two Sisters
booked a +60% revenue lift after their first causal read (2024)
No pixel
no SDK, no data team needed
€99
full refund if it doesn't move a decision
Dutch DTC teams switched to causal attribution. Not because we're great salespeople. Because once you see which channels actually make you money, you can't unsee it.
Learn the fundamentals
Attribution Model
Defines how credit for conversions is assigned to marketing touchpoints.
ROAS
Measures the revenue earned for every dollar spent on advertising.
Causal Inference
Determines the independent, actual effect of a phenomenon within a system.
Correlation
A statistical measure showing a relationship between variables; it does not imply causation.
Data-Driven Attribution
Platform ML model that distributes conversion credit based on observed data patterns.
Multi-Touch Attribution
Assigns credit to multiple marketing touchpoints across the customer journey.
Marketing Funnel
Describes the customer's journey from initial awareness to purchase.
Customer Acquisition Cost
The cost to convince a consumer to buy a product or service.
Customer Lifetime Value
Predicts the net profit from a customer's entire future relationship.
Cross Channel Attribution
Determines how different marketing channels contribute to customer conversions.
First-Touch Attribution
Gives 100% of conversion credit to the first marketing touchpoint.
Compare attribution tools
Causality Engine vs Triple Whale
Feature comparison for e-commerce attribution
Causality Engine vs Northbeam
Enterprise attribution comparison
Causality Engine vs Rockerbox
Multi-channel attribution comparison
Causality Engine vs Hyros
Ad tracking attribution comparison
Causality Engine vs Wicked Reports
Revenue attribution comparison
Further reading
Why Beauty Brands Need Causal Attribution
Deep dive into beauty-specific attribution challenges
TikTok Ads Attribution for Shopify
Track the full TikTok customer journey
The True Cost of Bad Attribution
How much are you wasting on misattributed spend?
Why Your Marketing Data Is Lying
The correlation vs causation problem explained
Questions TikTok-first founders actually ask
Is TikTok Shop actually making me money or just going viral?
Views are not revenue. Upload your free Google Analytics report and get a causal read in minutes. Formula: Incremental Sales = Revenue with TikTok - Revenue without TikTok. If TikTok starts journeys other channels close, last-click is hiding it.
How do I prove to finance that creator marketing works?
Dashboards log creators as zero because there is no pixel on word of mouth. Causal analysis measures lift: Revenue with creators - Revenue without creators. One confidence-scored number finance understands.
Do I need a data team or a pixel to run this?
No. Grab your free Google Analytics report (20 seconds), upload it (5 seconds), first causal read in 5-10 minutes at €99. No code, no pixels, nothing to install.
Why does Meta claim every sale when my brand grew on TikTok?
Meta usually has the last click, and last-click gives the closer 100% of the credit. TikTok opens journeys Meta closes days later. Causal attribution splits credit by what actually moved the sale.
What does it cost to find out?
€99 one-time for your first causal read, minutes after upload. Pro is €299/month with unlimited uploads, cancel any time.
Skip the queue
Get called in 60 seconds, or take the TikTok-first brands playbook.
Leave your number and we call you back within a minute to qualify fit and book a demo. Email-only is fine too: one short note a week, tuned to TikTok-first brands, unsubscribe in one click.
Or, if you are ready, book a 20-min call.
Find out if TikTok is making you money
One free Google Analytics report in. One causal read out. Minutes, not months.
No code. No pixels. Cancel any time.
Real reports from TikTok-first brandss
See what causal attribution looks like for a TikTok-first brands.
Anonymised reports from TikTok-first brandss already using Causality Engine. Per-channel incremental ROAS with confidence intervals.
Browse Tiktok-First-Brands reports in the Attribution Library →