TV & CTV Measurement
TV measurement is the practice of quantifying who saw a television ad and what it achieved, evolving from panel-based ratings toward impression-level, cross-platform measurement across linear and streaming.
Key takeaways
- Legacy TV measurement used statistical panels to estimate ratings.
- CTV enables impression-level, census measurement of actual delivery.
- ACR (automatic content recognition) identifies what's on screen.
- Cross-platform reach and frequency is the industry's central unsolved problem.
Panels to impressions
Traditional TV ratings extrapolate from a statistical panel of metered homes. CTV, being digital, can count actual impressions delivered "” a census rather than a sample. The industry is mid-transition, with new currencies competing to replace or supplement legacy panel-based ratings.
ACR and cross-platform reach
Automatic content recognition (ACR) on smart TVs identifies the content and ads on screen, feeding both targeting and measurement. The hardest remaining problem is cross-platform reach and frequency: deduplicating a viewer seen across linear, multiple CTV apps, and mobile, which no single identifier spans.
| Legacy method | Statistical panels / ratings |
|---|---|
| CTV method | Impression-level census |
| Key signal | ACR (automatic content recognition) |
| Open problem | Cross-platform dedup of reach & frequency |
Frequently asked questions
What is ACR in TV measurement?
Automatic content recognition identifies what content and ads appear on a smart TV screen, providing granular data for both measurement and targeting.
Why is cross-platform TV measurement so hard?
Because a single viewer appears across linear, several CTV apps, and mobile, with no common identifier tying those exposures together, deduplicating true reach and frequency is genuinely difficult.