
Incrementality
Part of CTV campaign measurement
Comparing completed views with business outcomes
See what a CTV completion rate can establish, how to compare it with enquiries or sales, and which conclusions need stronger evidence.
Put completed CTV views beside business outcomes, each with its own definition. A completion records playback to an endpoint; an enquiry or sale records a later action. A high completion rate does not prove a strong business result. A low rate cannot by itself explain a weak one.
Completed CTV Views vs. Business Outcomes: Key Differences
- Definition
- Playback to endpoint (completion); later action like purchase or enquiry (outcome)
- Measurement Timing
- During ad playback; after user interaction with business system
- Evidence Type
- Playback event (not attentive viewing or brand recall)
- Use in Decision-Making
- Supports delivery verification; not proof of incremental business impact
Define the completion rate
A completion is a playback event recorded under the supplier’s stated method. Ask what endpoint it represents and how the event is captured.
A supplier’s rate still needs a denominator. Ask for starts, completions, the formula, ad lengths, and any differences between publisher, player or ad-serving reports. Compare similar placements and periods before drawing a conclusion.
A completion describes a playback event. It does not show attentive viewing, brand recall or later action. Ask which playback conditions the measurement method can detect for the inventory in the report.
Key Metrics for CTV Campaign Evaluation
- Completion Rate Formula
- Completions ÷ Starts
- Ad Length Consideration
- Must be consistent across comparisons
- Reporting Variability
- Publisher, player and ad server reports may differ
- Attribution Limitation
- Not proof of incremental sales even when linked
Define the outcome separately
Choose the business event before launch. A retailer might track recorded purchases; a service business might track submitted enquiries. Specify what qualifies, when it is counted and whether it can be recorded consistently.
If the campaign aims to build awareness, an appropriately designed brand measure may be closer to the objective than immediate sales. Completion data, brand effects and sales effects are different forms of evidence.
For each comparable period or inventory group, show the delivery context, the completion count and denominator, and the agreed outcome. Record creative length, audience, publisher mix and other marketing alongside them. Differences in those conditions can affect interpretation.
Do not divide total sales by completed views and label the quotient a conversion rate. The counts may come from different populations and recording systems. Even where an attribution method connects eligible exposures with actions, its result is an attributed measure, not proof of incremental sales.
Use differences to decide what to investigate
When completion changes but the business outcome does not, check whether creative length, format or event handling changed. When the outcome changes but completion does not, inspect the offer, response path, audience and other activity. These checks can identify plausible explanations; dashboard trends alone cannot isolate a cause.
If outcomes are the primary goal, first confirm that the intended TV inventory, audience and dates ran. Then check that the response path and business event were recorded as planned. If those checks are sound but impact remains unclear, use a suitable comparison design before making a causal claim.
A final report can state the completion rate under its specified method, then state what happened to the agreed outcome under its own method. Name the uncertainty between them. Completions remain useful playback evidence while the campaign decision stays tied to its purpose.
How to Evaluate CTV Campaigns Using Completion and Outcome Data
- Confirm campaign deliveryVerify intended TV inventory, audience, dates and placements ran as planned
- Validate response path and event trackingCheck that enquiries or purchases were recorded correctly and consistently
- Compare completion rate and outcome trendsLook for discrepancies—e.g., high completions but low conversions
- Investigate contributing factorsCheck creative length, offer quality, audience targeting, and response path usability
- Use comparison design before claiming impactApply a controlled test (e.g., holdout group) to isolate causal effects



