The Media Rating Council (MRC) counts a display ad impression as viewable when at least 50 percent of its pixels are in view for one continuous second, and a video ad impression when the same share is in view for two continuous seconds, as stated in its August 18, 2015 Version 2.0 announcement. That is a counting rule. It says nothing about whether anyone paid attention.
What is the core rule?
The rule has two tests, applied in order. According to the Version 1.0 text, the pixel test comes first: at least 50 percent of the ad’s pixels must sit on an in-focus browser tab, in the visible part of the page. Only then does the clock start, and the time must run without a break. The same text adds that sub-second glimpses, or glimpses under 50 percent, are not added together.
For video, the same pixel share must hold for two continuous seconds of playback. Those seconds need not be the first two seconds of the ad, according to the same text.
| Format | Pixels in view | Continuous time | Source |
|---|---|---|---|
| Display | 50 percent | 1 second | MRC, 2015 |
| Video | 50 percent | 2 seconds | MRC, 2015 |
| Display, 242,500 pixels or more | 30 percent | 1 second | MRC, 2014 |
When does the large-ad exception apply?
Big formats struggle to get half their area on screen. Version 1.0, final on June 30, 2014, therefore lets a display ad of 242,500 pixels or more (the area of a 970 by 250 ad) count as viewable at 30 percent of pixels for one continuous second. The condition is disclosure: a measurer that uses 30 percent instead of 50 percent should fully disclose that to data users.
The provision appears in the Version 1.0 text. The Version 2.0 summary of key changes does not list it, but confirm it in the current full guideline before you write it into a contract.
What changed in Version 2.0, and what does the standard leave out?
Version 2.0 is dated August 18, 2015 and updates Version 1.0 of June 30, 2014. It followed reconciliation projects that looked for the causes of count differences among accredited vendors. The announcement says the baseline standards for desktop viewable impressions stayed the same.
The scope is narrower than the headline numbers suggest. The summary says these guidelines are intended for desktop advertising, that the MRC issued separate interim guidance for mobile, and that desktop and mobile counts should be reported separately. The 2014 text also said in-app impressions were generally assumed viewable under 2013 mobile app guidelines. That is 2014 wording, not a statement about today.
How do the reported numbers fit together?
A report should split served impressions into three buckets: viewable, non-viewable and undetermined (for example, when a cross-domain frame blocks the check). The Version 1.0 text defines three metrics:
- Measured rate: viewable plus non-viewable, divided by served impressions.
- Viewable rate: viewable divided by viewable plus non-viewable.
- Impression distribution: each bucket’s share of served impressions.
In the guideline’s own example of 1,000 served impressions, with 300 viewable, 200 non-viewable and 500 undetermined, the measured rate is 50 percent and the viewable rate is 60 percent. The same text says the 30 percent viewable share of the distribution should not be called the viewable rate.
What does this mean for a buyer?
A threshold is a definition. Accreditation is separate: the MRC describes it as applying to measurement services that disclose their methods, meet MRC standards and submit to MRC-designed audits. Vectravia does not claim MRC accreditation, and this article describes the standard only.
Signal Reports reports viewability against these thresholds, 50 percent of pixels for one second on display and two seconds on video, and labels measured and modeled numbers apart. When you compare any vendor, ask which guideline version it follows, how many impressions were undetermined, how it uses the large-ad exception, and whether the pixel test uses the ad or the video player. The Version 1.0 text asks for disclosure when player pixels replace ad pixels.