Viewability Rate: Calculate, Diagnose and Improve Visible Delivery
Calculate and interpret viewability rate with a documented denominator, measurement standard, placement breakdown and outcome-level quality review.
How is ad viewability measured?
Ad viewability measurement classifies whether a measurable impression met the area and continuous-time conditions of a named standard for its format. Viewability describes an opportunity to see within the supported environment; it does not prove attention, comprehension, intent or outcome.
This page focuses on the measurement method used by a buyer or analyst. It distinguishes eligible, measurable, viewable, non-viewable and non-measurable populations; explains denominator and coverage risk; and reconciles vendor reports without treating one provider's field as universal truth.
Measurement references checked 2026-08-10: MRC guidelines, IAB Tech Lab Open Measurement material and current Google Active View documentation supply the standards and platform examples used below. FTC and WCAG references define truth and accessibility boundaries.
1. Name the standard, format and provider
Record the measurement provider, methodology version, accreditation or certification scope where applicable, ad format, device environment and reporting product. Display, video, large-format and application inventory can use different eligibility and viewability rules.
Do not shorten every provider field to MRC viewability. State whether the result is provider-reported, independently verified or derived from exported fields. A standard supplies criteria; the implementation determines which impressions can be measured against them.
2. Preserve the population funnel
Keep total or served, eligible, measurable, viewable, non-viewable and non-measurable impressions as separate fields. Google Ad Manager describes measurable impressions as a subset of eligible impressions and viewable impressions as a subset of measurable impressions.
A missing or non-measurable result is not evidence of failure to meet the viewability threshold. It is a coverage state. Report every bucket needed to reconstruct the rate and avoid removing inconvenient inventory from the denominator without disclosure.
3. Apply area and time rules to the correct creative state
The measurement client must identify the creative geometry, browser or application viewport, continuous elapsed time and relevant playback or expansion state. Current Google methodology describes display and video rules based on visible area and uninterrupted duration under supported implementations.
Use the exact threshold recorded by the provider and standard version. Do not claim that a partial, interrupted or background state qualified when the measurement record does not support it. Large-format and expandable treatments require their applicable rules.
Preserve whether the required duration was continuous and which event reset the clock. Several short appearances should not be combined unless the governing method explicitly permits that treatment. The report should also state whether a second qualifying period can create another viewable event for the same served impression.
4. Distinguish measurable rate from viewable rate
Measurable rate describes how much eligible delivery the technology could assess. Viewable rate describes how much measurable delivery met the viewability rule. A strong viewable rate on low measurement coverage can provide an incomplete campaign picture.
Publish the numerator, denominator and coverage beside each percentage. Avoid a generic viewability score that mixes measurable and total impressions differently across reports. Name whether a distribution field uses total or measurable delivery.
5. Validate web measurement implementation
Inspect tag execution, iframe context, browser support, viewport geometry, tab state, scrolling, resizing, asset render and measurement callbacks. Current Google methodology describes use of browser signals such as Intersection Observer in supported web environments.
Test production placements rather than a standalone creative preview. Consent, content-security policy, script failure, sandboxing, cross-domain configuration and late render can change measurability. Avoid adding duplicate verification scripts that increase latency without a declared decision use.
Viewability population contract
Every percentage needs a named population and coverage rule.
| Population | Meaning | Valid calculation use | Unsafe inference |
|---|---|---|---|
| Eligible | Delivery supported for assessment | Measurable-rate denominator | Successfully measured |
| Measurable | Technology captured required signals | Viewable-rate denominator | Viewable |
| Viewable | Met applicable area and time rule | Qualified opportunity count | Human attention |
| Non-viewable | Measured but did not qualify | Placement diagnosis | Invalid traffic |
| Non-measurable | Viewability state unavailable | Coverage reporting | Failed threshold |
6. Validate mobile-app and video support
For application inventory, record SDK, OMID declaration, certified partner context, application lifecycle and supported playback signals. For video, preserve player state, continuous playback time, visibility and format eligibility.
Do not compare a web display field with an in-app video field solely because both are called viewable. Confirm whether the provider reports accredited and non-accredited inventory separately and how unsupported environments enter the denominator.
7. Treat occlusion and background state explicitly
Document whether the implementation can detect clipped content, inactive tabs, picture-in-picture, application backgrounding, overlays and operating-system windows. Coverage can vary by platform and measurement technology.
When occlusion cannot be determined, report the limitation rather than assuming the creative was unobstructed. Reproduce suspicious placements on representative devices and preserve the observed state with provider documentation and timestamps.
8. Filter invalid traffic before rate interpretation
Confirm where invalid-traffic filtering occurs in the provider pipeline and whether total, measurable and viewable fields share the filtered population. Different timing or methods can create discrepancies between buying, serving and verification reports.
Keep provider filtering and independent anomaly review separate. Do not call a non-viewable impression invalid or an unmeasurable impression fraudulent. These fields answer different questions and require different evidence and remediation.
9. Reconcile measurement vendors by event scope
Align campaign, creative, placement, format, device, date, time zone, served-event definition, filtering and supported inventory before comparing vendors. Preserve raw exports because UI totals can change after processing or methodology updates.
Set a material discrepancy threshold and escalation route in advance. A difference can reflect coverage, timing, sampling, invalid traffic, iframe support or creative geometry. Do not average incompatible viewable counts into a synthetic truth.
Build the comparison from joined impression identifiers where contracts and privacy controls permit; otherwise reconcile at the narrowest common aggregate. Report match coverage, unmatched populations and processing delay. A close top-line total can still hide opposite placement-level errors that cancel one another.
10. Segment by placement and environment
Break out viewability and measurability by source, placement, page position, format, device, browser, application and creative size where volume allows. An account-wide rate can hide one unmeasurable environment or a high-volume low-viewability placement.
Retain volume beside percentages and apply minimum evidence. A small placement can display an extreme rate by chance or because reporting thresholds removed events. Prioritize repeatable, material patterns rather than ranking noise.
11. Diagnose non-viewable delivery without assuming attention
Review render timing, placement position, scroll behavior, creative size, page depth, playback state and source context. The remedy should address the observed mechanism, such as late rendering or a placement rarely entering the viewport.
A viewable impression still does not establish that a person looked at or understood the ad. Keep attention studies, interaction and accepted outcomes separate. Do not turn viewability eligibility into a claim of verified viewing.
Viewability discrepancy diagnostic
Align implementation before interpreting a vendor difference.
| Observed gap | Possible mechanism | Evidence needed | Response |
|---|---|---|---|
| Measurability drops | Tag or environment support | Execution and coverage | Repair supported path |
| Vendors report different rates | Population or threshold | Method and event exports | Report separately |
| Desktop and app diverge | SDK and lifecycle | Environment-specific logs | Validate implementation |
| Rate rises, coverage falls | Narrow measurable subset | Eligible and measurable counts | Qualify improvement |
| Viewability rises, value flat | Opportunity not outcome | Accepted records and maturity | Hold causal claim |
12. Calculate cost with the qualified denominator
Relate measurable cost and viewable CPM only to the fields defined by the provider. State whether cost covers all served delivery or a qualified subset and whether non-measurable spend remains included elsewhere.
Compare served CPM, measurable coverage, viewable CPM, reach, frequency and accepted outcomes together. A higher viewable CPM can accompany stronger qualified delivery, while a low rate can reflect cheap inventory that rarely satisfies the measurement rule.
13. Connect viewability with campaign outcomes
Join campaign and creative viewability evidence with destination behavior and accepted outcomes under the same time, attribution and maturity contract. Use viewability to explain opportunity quality, not to replace the declared business result.
Run controlled tests when a viewability intervention changes material spend or inventory. Preserve source mix and creative conditions. A correlation between viewable rate and outcome does not prove that the measured exposure caused the change.
14. Protect performance and accessibility during measurement
Use measurement code that is necessary, supported and monitored. Avoid redundant tags, synchronous loading and scripts added only to satisfy a dashboard. Measure their effect on render, interaction and layout stability in representative production conditions.
Viewability optimization must not obscure content, trap focus, create motion hazards or make the destination inaccessible. WCAG compliance and user experience remain release conditions even when a placement could improve its measured opportunity.
15. Maintain a viewability method record
Record provider, methodology version, standard, accreditation scope, format, environment, event funnel, thresholds, filtering, coverage, source dimensions, costs, discrepancies, outcome join and known limitations. Link every conclusion with the exact exports used.
Revalidate after SDK, tag, browser, creative, supply, format or provider changes. Publish a dated break in series when methodology changes materially. Do not compare historical and current rates as one continuous measure without qualification.
16. Use FroggyAds delivery in a qualified viewability review
Use FroggyAds campaign, format, source, bid, budget and delivery evidence as available within the self-serve platform. Add viewability fields only from supported platform or verification reporting and retain their provider-specific definitions.
Do not infer viewability for formats or environments where the necessary measurement is unavailable. Reconcile qualified exposure with frequency, destination behavior, cost and the advertiser's accepted outcomes before changing budget or supply.
Questions about viewability measurement
What is a viewable ad impression?
It is a measurable impression that met the applicable visible-area and continuous-time rule for its named format and methodology.
Is viewability the same as attention?
No. It establishes a qualified opportunity to see under the measurement method, not verified human attention or comprehension.
What is measurable rate?
It is the share of eligible impressions for which the measurement technology could capture the required viewability signals.
What is viewable rate?
It is the share of measurable impressions that satisfied the applicable viewability criteria under the provider's method.
Are non-measurable impressions non-viewable?
No. Their viewability state could not be determined, so they require separate coverage reporting.
Why do viewability vendors disagree?
Coverage, event scope, filtering, timing, format support, iframe behavior and methodology can differ.
Can viewability be compared across formats?
Only with explicit format-specific rules, populations and environments; one unqualified rate can hide material differences.
Does higher viewability guarantee better outcomes?
No. Creative meaning, audience, source, frequency, destination and mature accepted value still determine campaign performance.
When should a viewability series be restated?
Restate or mark a break after material changes to provider, methodology, SDK, tag, filtering, format support or population definition.
How should FroggyAds viewability be reported?
Use only supported viewability evidence with its provider definition and reconcile it with FroggyAds delivery, cost and accepted outcomes.
Primary references for viewability methods and standards
- MRC viewable ad impression measurement guidelines
- MRC digital video served-impression guidelines
- Google Active View methodology
- Google Ads Active View reporting metrics
- IAB Tech Lab Open Measurement SDK
- Google Ad Manager Active View calculations
- US FTC truth-in-advertising guidance
- W3C Web Content Accessibility Guidelines 2.2
Qualify viewability before changing campaign supply
Use FroggyAds after provider, event populations, format rules, coverage, cost and accepted-outcome evidence are aligned.
Create My Free Account