A 70% viewability rate needs a second number

A placement reports 70% viewability. Before calling it strong or weak, ask how much of its delivery could actually be measured.

Viewability describes whether an impression met the applicable opportunity-to-see criteria. Measurability describes whether the system obtained the information needed to make that assessment. Those are separate questions.

Google’s Active View reporting guidance distinguishes measurable impressions from viewable impressions and calculates the viewable percentage using measurable delivery. Use the definitions of the measurement system on your plan rather than assuming every report has the same coverage.

Reconstruct the denominator

Suppose a fictional placement has 100,000 served impressions, of which 60,000 are measurable and 42,000 are viewable.

Its viewability rate is 42,000 divided by 60,000, or 70%. Its measurement coverage relative to served impressions is 60%. Dividing 42,000 by all 100,000 produces 42%, which answers a different question.

The remaining 40,000 impressions are not automatically non-viewable. Their viewability is unknown under this measurement. Labeling them failures would overstate what the data establishes; ignoring their existence would hide a coverage limitation.

The viewability calculator helps with the rate. Keep the source counts beside the result.

Look for changes in coverage

Imagine the next report shows 80% viewability but only 20,000 measurable impressions. That could reflect a better experience among the measured delivery, a change in the measured inventory, or a measurement problem elsewhere.

Compare the same placements, devices, formats, and measurement implementation. Check whether new inventory supports the same verification method. Request an explanation for large coverage changes before describing the higher percentage as a campaign improvement.

Where available, examine the split between measurable-but-not-viewable delivery and unmeasurable delivery. These categories lead to different investigations.

Connect the number to the page experience

For a web placement, review where the slot appears, how it loads, and whether users scroll to it. For video or app inventory, confirm the relevant environment and measurement support.

Do not substitute a screenshot of one successful rendering for campaign-wide measurement. A preview establishes that an ad can appear in a particular test; it does not establish the proportion of delivered impressions that met the criteria.

Likewise, viewability is not proof that a person noticed, understood, or acted on the message. It should sit beside the objective and other appropriate evidence.

Report the limits without burying the result

A useful summary is: “70% of measurable impressions were viewable; 60% of served impressions were measurable under this provider’s method.”

Add the period, source, format, and any major exclusions. If comparing partners, align those definitions before ranking them.

When a placement changes, inspect both counts again. The next decision should follow a clear diagnosis: improve the viewing opportunity, repair measurement coverage, or investigate a change in delivery mix. A single percentage cannot choose among those actions.