Video completion rate (VCR) is the percentage of video ad impressions where the viewer watched the ad to the end — the final frame, or at minimum the 97% threshold used by most measurement platforms. In CTV, VCR is the primary campaign performance metric because CTV ads are almost always non-skippable: a viewer cannot click away, so completion rate reflects genuine attention and ad exposure, not just whether an impression was served.
How VCR is calculated
VCR = (completed views ÷ total impressions) × 100
A completed view is defined by the IAB as reaching the 97% mark of the ad duration. So a 30-second ad is "completed" when the viewer has watched at least 29.1 seconds. Most DSPs and SSPs use this definition, though some publishers report completion at 100% — always confirm which definition is being used when comparing benchmarks.
VCR sits at the end of the quartile reporting chain: 25% (first quartile), 50% (midpoint), 75% (third quartile), 100% (completion). A sharp drop between 75% and 100% indicates viewers are abandoning just before the end — common when the creative's message lands early and the final seconds feel like dead air.
Why VCR matters more in CTV than in mobile or desktop video
In mobile and desktop video, skip rates are a confounding variable. A 60% VCR on a skippable YouTube pre-roll means something very different from a 60% VCR on a non-skippable CTV ad — the latter represents a genuine measurement problem (buffering, mid-roll abandonment, app crash), not viewer choice.
CTV VCR should be high by default. Non-skippable, lean-back viewing on a 10-foot screen with no competing tabs or notifications is the highest-attention environment in digital advertising. If your CTV VCR is below 80%, something is technically wrong — the most common causes are buffering from CSAI (client-side ad insertion), poor VAST tag implementation, or the ad server timing out before the creative loads.
India CTV VCR benchmarks
Based on industry data from India CTV campaigns:
- Premium publisher, SSAI-delivered: 90–97% VCR. JioHotstar, SonyLIV with server-side ad insertion consistently deliver near-full completion rates on non-skippable inventory.
- Mid-tier publisher, CSAI: 75–88% VCR. Client-side ad insertion introduces buffering and timeout failures that reduce measured completions even when the viewer was present.
- FAST channel / long-tail CTV: 70–85% VCR. Lower engagement content and less robust ad tech infrastructure.
- Open auction, mixed publishers: 78–85% average VCR. Wide variance depending on the publisher mix that clears your bids.
A VCR below 75% on a CTV campaign should trigger an audit: check for CSAI vs SSAI, VAST error rates, and whether your DSP is reporting measured completions or estimated completions.
VAST errors and VCR inflation
VCR figures can be misleading in one direction: inflation from VAST errors. When a VAST tag fails to load, the impression may not be counted at all — meaning only the successfully delivered ads appear in the denominator. A publisher with 20% VAST error rate will show artificially high VCR on the impressions that did deliver. Always look at VAST error rate alongside VCR. Error rates above 5% indicate a technical problem that is reducing both reach and the reliability of VCR as a metric.
VCR as a publisher quality signal
For buyers, VCR by publisher is a proxy for inventory quality. Publishers with consistently high VCR (90%+) have better technical infrastructure and more engaged audiences. When evaluating new CTV publishers for PMP deals, ask for historical VCR data by content type. A publisher who cannot provide this data either doesn't have adequate measurement in place or is hiding poor performance — both are reasons to negotiate a lower floor or test with a small budget before committing.