Back to all blog posts

Viewability measurement – here’s how it’s done

By Asaf Shamly | March 24, 2021

viewability measurment

Ever since ad viewability became a currency in the digital media world some five years ago, publishers have been at a disadvantage. While advertisers are arming themselves with different tools and technologies to improve their bottom line, publishers are in a constant struggle to hit campaign goals and meet industry viewability standards. They keep losing money on wasted ad impressions that were deemed unviewed.

The playfield is tilted in the advertisers’ favor.

This post aims to empower publishers with knowledge that will hopefully improve their position in a negotiation and get more impressions deemed as viewed.

I’ll shed light on how viewability is measured, what the MRC guidelines state, and where things sometimes get tricky.

Let’s get to it.

What is ad viewability?

Both the Interactive Advertising Bureau (IAB) and the Media Rating Council (MRC) define a viewable impression as one that:

  • Has at least 50% of the ad’s pixels on an in-focus browser tab, on the viewable space of the browser page
  • Meets the above requirement for at least one continuous second.

The guidelines above should be taken as a rule of thumb, with a few exceptions. For example, if an ad is super large (covering at least 242,500 pixels – an equivalent of a 970 x 250-pixel ad), at least 30% of it needs to be in view to be considered a viewable impression.

How is ad viewability measured?

Usually, ad viewability is measured by adding a Javascript tag to the ad. In some cases, the measurer may also add the tag to the ad container in which the ad appears, such as the I-Frame. MRC says that, while it won’t speak against tagging the container, it doesn’t fully recommend it:

“Viewability measurement based on the ad container involves an inference that that ad in fact appeared within the container in its intended format,” the guidelines read.

“While measurement based on the ad itself is generally preferable whenever possible, ad container-based measurement is also acceptable under these guidelines, but it should be supported by evidence that viewability measurement based on the container rather than the ad does not result in material counting differences, or in inaccurate viewability determinations because of the mis-sizing or scaling of ads that appear within the container.”

What are the most common ad viewability measurement tools out there?

There are a handful of tools which are generally perceived as the industry leaders and choosing between them is, more or less, personal preference.

Among the most popular tools are MOAT, DoubleVerify, Integral Ad Science (IAS), and Google’s Active View (all of which are MRC accredited, by the way).

Many publishers will gravitate towards MOAT, as it allegedly measures 100% of the ad impressions (I know this was supposed to be a given, but apparently – it’s not) and has a sound Invalid Traffic (IVT) detection mechanisms.

They all offer similar features, from proprietary tools that enable publishers to evaluate and segment their ad inventory through viewability, to ad viewability data on campaign, domain, ad placement, and many other levels. Google’s Active View is unique in the sense that its focus is on Google’s Display Network and YouTube.

Where do things get tricky?


The discrepancy issue – comparing apples to oranges

Keep in mind that a change in the methodology each tool uses could also mean different results compared to the advertiser’s. You could potentially end up with mismatching numbers. At the end of the day, that could mean less (or more!) profit, so it would be wise to at least know how your advertisers measure viewability and which tools they use.

The most common metric that will potentially be tracked differently is milliseconds

Let’s recap for a second (pun intended!) – an ad is considered viewed if half of it was in the visitor’s view for at least one full second. But how do you measure that one second? MRC’s guidelines state that you should be measuring in 100 millisecond batches, which means that an ad is considered viewed if it was in the visitor’s uninterrupted line of sight for 10 consecutive 100 millisecond batches. It will, to an extent, allow for slight deviations from this norm, if the measurer can prove that the results won’t differ greatly.

“If the measurer can empirically validate that its polling for viewable display impressions at a less frequent interval than every 100 milliseconds (but no less often than at every 200 milliseconds) will result in no material differences in its viewable impression counts, this less frequent interval is allowable for display impression measurement,” MRC explains in its guidelines.

“If a measurer chooses to poll for display viewability at a less frequent interval than 100 milliseconds, the support for this approach must be revalidated on at least an annual basis, and this practice must be prominently disclosed.”

MOAT is an example of a digital ad analytics tool that would rather do their own thing. They claim to be checking the visibility of the ad every 200 milliseconds and will wait for five consecutive positive checks before deeming the ad impression as viewable. They also say that an internal investigation led to the conclusion that the 200ms methodology is “equivalent in accuracy”.

Other metrics that can potentially be tracked differently include:

  • expandable rich media ad units
  • browser sizing
  • ads appearing on pre-fetched or pre-rendered pages (MRC did warn about this in its guidelines)

In addition, how each tool manages IVT is an entirely different beast.

Since it’s not realistic to use all of the different tools just to be able to speak the same language as the advertisers, discrepancies are inevitable.

There’s no real solution here, but knowing what’s measured is a step in the right direction.

The DIY approach

A less common way to measure viewability is to roll up your sleeves and measure ad viewability yourself. This approach can, if anything, give you a better understanding of the mechanics behind the measurement.

That, paired with a third-party tool (for confirmation, for example), can give you a clearer and more detailed perspective on your inventory’s ad viewability and help you negotiate discrepancies with the advertiser to your favor.

Knowledge is power

By knowing the MRC’s guidelines and understanding traps that may appear along the way (such as the milliseconds conundrum or how different tools tackle expandable rich media ads) – you can arm yourself with a better understanding of the actual performance of your ad inventory.

I hope this article clarified how ad viewability is measured and will help you make more informed decisions going forward.

Latest Articles