Multi-Touch Attribution for CTV Campaigns

Last-click gives CTV zero credit. Multi-touch reveals reality: 38% of impressions drove 63% of conversions. Use position-based, 14-30 day windows.

Last-click attribution gives CTV zero credit. Multi-touch attribution reveals CTV’s actual contribution. Understanding attribution models is essential for measuring CTV accurately.

The Last-Click Problem

1

CTV Exposure

Customer sees your CTV ad while streaming their favorite show.

2

Brand Search

Days later, customer searches your firm name on Google.

3

Click & Convert

Customer clicks your Google ad and fills out the form.

4

Wrong Credit

Under last-click: Google Ads gets 100%, CTV gets 0%.

This is obviously wrong. CTV drove the search. But last-click can’t see that.

How Multi-Touch Attribution Works

Multi-touch models distribute credit across touchpoints:

ModelHow Credit Is Assigned
Last-click100% to final touchpoint
First-click100% to first touchpoint
LinearEqual credit to all touchpoints
Time-decayMore credit to recent touchpoints
Position-based40% first, 40% last, 20% middle
Data-drivenAlgorithmic based on actual conversion patterns

Linear Attribution Example

Journey: CTV → Organic search → Paid search → Conversion

  • CTV: 33.3%
  • Organic: 33.3%
  • Paid search: 33.3%

Position-Based Example

Same journey:

  • CTV: 40% (first touch)
  • Organic: 20% (middle)
  • Paid search: 40% (last touch)

Time-Decay Example

  • CTV: 15% (earliest)
  • Organic: 25%
  • Paid search: 60% (most recent)

Why Multi-Touch Matters for CTV

CTV is almost always an upper-funnel touchpoint. It creates awareness that other channels capture. Under last-click:

  • CTV looks like it generates zero leads
  • Budget gets cut
  • Performance actually declines (because CTV was driving search)

With multi-touch attribution:

  • CTV’s contribution is visible
  • Budget decisions are informed
  • System optimization becomes possible

CTV accounted for 38% of impressions but 63% of conversions when proper attribution was applied.

Choosing an Attribution Model

Option A

Option B

Position-based gives appropriate credit to CTV as the awareness driver while still crediting conversion channels.

Implementing Multi-Touch Attribution

Option 1: Platform-Native

Many CTV platforms include attribution:

  • MNTN tracks verified visits through conversion
  • Tatari provides multi-touch models
  • Some integrate with GA4

Option 2: Third-Party Attribution

Dedicated attribution platforms:

  • Rockerbox
  • Measured
  • Northbeam
  • Triple Whale

These aggregate data across channels for unified attribution.

Option 3: GA4 Attribution

Google Analytics 4 offers:

  • Data-driven attribution (default)
  • Multiple model comparison
  • Cross-channel visibility

Requires proper UTM tagging and integration.

Attribution Windows for CTV

How long should CTV get credit?

ScenarioRecommended Window
Direct response7-14 days
Consideration purchase14-21 days
Legal services14-30 days
High-value services21-30 days

PI legal services are consideration purchases. Someone sees your ad, has an accident weeks later, then searches for you. Attribution windows need to capture this.

Cross-Device Attribution

CTV plays on TVs. Conversions happen on phones and computers.

The challenge:

  • Ad seen on living room TV
  • Search happens on phone
  • Form filled on laptop
  • How do you connect them?

Solutions:

  • Household-level matching (same IP)
  • Device graphs (probabilistic matching)
  • Login-based matching (deterministic)

Platform match rates of 90%+ make cross-device attribution increasingly reliable.

Reading Attribution Reports

Key Metrics to Track

Assisted conversions: Conversions where CTV was in the path but not the last touch

Conversion paths: Common journeys (CTV → Search → Conversion)

Attribution credit by channel: How credit distributes across your mix

Model comparison: How credit shifts between models

What Good Looks Like

MetricHealthy Sign
CTV assisted conversionsGrowing with spend
CTV → branded search pathsCommon pattern
CTV attribution credit15-30% of conversions
Model varianceCTV credit higher in position-based than last-click

Common Attribution Mistakes

❌ Mistakes to Avoid

  • − Using only last-click. Cuts CTV because it 'doesn't convert directly'
  • − Over-crediting CTV. First-click or long windows inflate CTV credit
  • − Ignoring offline. Phone calls matter for law firms, missing them misses conversions
  • − Not testing models. Large differences between models suggest measurement issues

Attribution and Budget Decisions

Attribution informs but doesn’t dictate:

What attribution tells you:

  • Relative channel contribution
  • Journey patterns
  • Where to investigate

What attribution doesn’t tell you:

  • Absolute ROI
  • Optimal budget levels
  • Creative quality impact

Use attribution as input to decisions, not as the decision.

For complete measurement strategy, see the CTV attribution guide.

References