Track delivery, audience coverage, exposure quality, attention or brand response, and business outcomes—but keep their meanings separate. An impression shows that an ad was counted as delivered; it does not, by itself, prove that a person noticed it or that the campaign caused a sale. For campaigns spanning linear TV, connected TV (CTV), and streaming video, compare reports only after checking audience definitions, deduplication, co-viewing, impression rules, measurement coverage, and outcome methods.
Build a scorecard around five questions
Use a consistent set of measures, then break results out by channel, publisher or platform, placement, geography, device, and creative where those details are available. The table is a starting framework, not a claim that every seller supports every metric.
As an Amazon Associate I earn from qualifying purchases.
| Question | Metrics to track | What to check before interpreting them |
|---|---|---|
| Was the ad delivered, and at what cost? | Impressions, spend, and CPM | What event counts as an impression, what is excluded, and whether invalid-traffic controls apply. |
| How many people in the intended audience did the campaign reach, and how often? | Target-audience reach, total reach, average frequency, frequency distribution, and—across channels—unduplicated reach and overall frequency | Population, geography, time period, person- or household-based estimate, identity and deduplication method, and treatment of co-viewing. |
| How much of the ad was measured as viewable or completed? | Video quartiles and completion rate, measurable impressions, viewable impressions, viewable rate, and non-measurable share where available | The viewability definition, denominator, device and publisher coverage, and whether completion is being used as a proxy. |
| Did the campaign attract attention or change brand response? | Defined exposure and engagement signals; brand lift, recall, or another survey outcome when a sound study design is available | Provider-specific definitions, study population and dates, sample, uncertainty, and the exact outcome measured. |
| Did it contribute to a business result? | Advertiser-defined conversions, conversion volume, cost per result, and incremental lift where a credible causal design is available | Event definition, attribution window, identity matching, modeled components, and whether the result is attributed or demonstrated to be incremental. |
Measure delivery without mistaking it for impact
Start with impressions, spend, and CPM for each relevant channel and available breakdown. An impression is delivery evidence, not proof of a unique person, attention, viewability, or business impact. The IAB digital video guidelines cover online, mobile, and OTT video and audio activity, including streaming-player impression counting; the counted event and exclusions still need to be clear in each report.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallOutdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchFor streaming inventory, ask whether the impression count excludes invalid traffic and playback that continues while a television is off. A raw served-impression total should not be treated as a count of people reached or viewable exposures. Definitions may also differ among web video, CTV, social video, and linear television.
#1 Best Overall
Separate reach, frequency, and GRPs
Reach estimates how many unique people—or, depending on the method, households—were exposed during a stated period. Frequency estimates the average number of exposures among those reached. Ask for the frequency distribution as well as the average: the same average can conceal very different patterns of repetition across individuals.
For a campaign spanning linear TV and streaming, request cross-platform unduplicated reach and overall frequency, along with the population, period, and deduplication approach. Find out how the method handles multiple people watching the same CTV screen. Google reporting definitions, for example, describe unique and co-viewed reach as estimates based on identifiers and aggregated sign-in data, not direct person-by-person counts.
GRPs combine reach and frequency, but do not reveal how exposures are distributed among individuals. Nielsen noted in its 2023 audience-measurement explainer that frequency control can be difficult across platforms and that TV advertisers often buy GRPs while individual-viewer frequency remains largely uncontrolled. Treat a GRP as a summary, not a substitute for cross-platform reach and frequency reporting.
Keep completion, viewability, and measurability distinct
Completion rate describes how often playback reaches the defined end point; quartiles show progress through the video. Neither automatically establishes that the ad was viewable. Viewability uses a separate measurement rule, and its coverage can vary by device, publisher, and measurement method.
Google Active View’s video definition counts an ad as viewable when at least 50% of its area is visible for at least two seconds. That is a Google platform definition, not a universal threshold or proof that every CTV impression can be measured this way. Google also distinguishes a non-measurable impression—where viewability could not be determined—from a non-viewable impression. Report the measurable denominator and non-measurable share so readers can tell what the rate represents.
Choose attention and brand metrics to fit the objective
Attention providers use different methods and scores, so do not assume that scores from different providers are directly comparable. IAB’s attention framework distinguishes ad-exposure signals, such as viewable time, share of screen, completion, and audibility, from user-engagement signals, such as clicks, skips, playback events, and volume changes. Interpret each signal according to its definition rather than combining unlike signals into an unexplained “attention” result.
Rank #3
- Used Book in Good Condition
Evaluate the media and content environment too: fraud controls, viewability, brand safety, suitability, contextual relevance, and geographic targeting affect how useful exposure signals are. For an awareness or persuasion campaign, add a defined brand-lift or recall measure when an exposed/control or pre/post design is credible. Report who was studied, the sample and field dates, the exact survey outcome, and its uncertainty. A modeled estimate is not a directly observed count.
Define business outcomes and incrementality
Before launch, specify the event that counts as a result for this advertiser: for example, a qualified visit, lead, purchase, or app action. Report conversion volume and cost per result, and distinguish attributed outcomes from incremental outcomes. Attribution assigns credit under a stated rule; incrementality asks whether the campaign caused additional results compared with a credible counterfactual.
Disclose the attribution window, identity-matching approach, modeled components, and any control group or other causal method. IAB Tech Lab describes ECAPI as a standard for communicating marketing-related events from advertisers’ systems to platforms and partners. Standardized event communication can help with data exchange, but event plumbing alone does not establish causal lift.
Rank #4
Compare reports and vendors on the same basis
Before comparing two campaigns, publishers, or measurement providers, check the underlying definitions and coverage—not just the headline totals. Request the following details in writing:
- Audience scope: Target population, geography, demographic basis, reporting period, and whether estimates are person- or household-oriented. Ask how panels and large-scale device data are combined and how viewing is represented.
- Deduplication and co-viewing: Whether exposures are deduplicated across devices, platforms, and linear TV, and how a shared CTV screen is treated.
- Impression and traffic rules: The counted event, exclusions, invalid-traffic controls, and whether definitions align across web video, CTV, social video, and linear television.
- Visibility coverage: Applicable viewability standard, measurable denominator, non-measurable share, and whether completion or progress events stand in for viewability.
- Outcome design: Event definitions, attribution windows, observed versus modeled results, control or incrementality method, and uncertainty.
- Independent assurance: MRC accreditation where applicable, standards followed, and independent verification of impression, invalid-traffic, and viewability methods.
- Reporting access: Available breakdowns, data latency, and access to underlying event or impression data where contractually and legally permitted.
These are useful vendor questions for CTV and streaming measurement:
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- Which signals do you use—such as SDKs, tags, or server-side logs—and how do device differences affect measurement?
- Do your estimates use deterministic signals, modeled or probabilistic methods, panel data, or a combination? How are panel estimates generalized?
- How do you count completion, and how does that differ from impression counting and viewability?
- How do you detect playback that continues when a television is off?
- Is the measurement platform MRC-accredited? If not, which standards does it follow, and has the method been independently verified?
IAB’s public summary of its December 2025 CTV guide identifies fragmented standards and inconsistent signal quality as ongoing challenges. IAB Tech Lab’s measurement page listed OM SDK v1.5 (June 2024) for third-party viewability and verification in web video, mobile apps, and CTV, and ECAPI v1.0 (January 2026) for standardized event communication when accessed. Check the current applicable version and implementation with the vendor; a listed standard does not guarantee identical coverage across sellers or devices.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
