Video Measurement and Attribution

Uniquely hard. Metrics hierarchy, platform-specific, brand lift, cross-screen, MMM, incrementality.

Why video measurement is hard

Video lives at intersection of brand and performance. View-through dominates. No click on CTV. Cross-screen identity stitching imperfect. Single-metric attribution misleads; triangulation essential.

Metrics hierarchy

• Reach: Unique households/people.
• Frequency: Exposures per person.
• Completion rate: % watched to end.
• View-through rate.
• Brand lift: Awareness, consideration, intent.
• Search lift: Brand search increase.
• Sales lift: Causal sales impact.
• CTR / conversion: Where applicable.

Platform measurement

• YouTube: Brand Lift, Conversion Lift, Search Lift studies native.
• Meta video: Brand Lift, Conversion Lift.
• TikTok: Brand Lift.
• CTV (DSP): Verification + Brand Lift partners.
• Netflix, Disney+: Limited measurement; black box.

Brand lift

• Survey-based: test exposed vs control.
• Awareness, consideration, intent, favorability.
• Statistically significant lift = ad-driven.
• Cohort sizing matters.
• Quarterly cadence for major campaigns.

Cross-screen

• Same household watches CTV + mobile + desktop.
• Identity stitching via household graphs (LiveRamp, ID5).
• ACR data on smart TVs.
• Cross-screen reach measurement difficult.
• Frequency uncapped across screens common waste.

MMM for video

• Aggregate video spend modeled against outcomes.
• Calibrate with incrementality tests.
• Bayesian frameworks include video as channel.
• Brand vs performance channels distinguished.

Incrementality

• Geo holdouts.
• User-level holdouts where DSP supports.
• Brand lift studies as proxy.
• Conversion lift where applicable.
• Quarterly testing rotation.

Advanced playbook

• Cross-screen frequency management.
• Brand lift studies on major campaigns.
• Incrementality testing quarterly.
• MMM annual rebuild.
• Triangulation reporting.
• Identity strategy across screens.
• Reach + frequency optimization.
• View-through with discount factor.
• Walled garden vs open measurement.
• Annual measurement methodology review.

Common mistakes

• Single platform attribution treated as truth.
• Brand lift skipped.
• Incrementality not tested.
• Cross-screen frequency uncontrolled.
• Identity stitching naive.
• View-through credit without discount.
• MMM not built.
• Walled garden inventory measurement ignored.
• CTV measured by digital KPIs.
• Triangulation absent.

Operating checklist

• Brand lift on major video campaigns
• Incrementality testing quarterly
• MMM annual rebuild
• Cross-screen frequency management
• Identity strategy across screens
• View-through with discount factor
• Walled garden inventory considered
• Triangulation reporting
• Annual measurement review
• Stakeholder education on limits

Part of the YouTube & Video Marketing series.