Viewability Examined

CURRENT STATE OF VIEWABILITY

Measuring whether an ad is actually seen

Viewability is an important metric for marketers to understand as it relates to the overall performance of a digital campaign. It reflects the amount of time that an ad unit is measured as “in view” after the ad serves. The current Media Rating Council’s (MRC) Standard of Viewability states that a “minimum of 50 percent of pixels are in view for one consecutive second for display and two consecutive seconds for video.”

Recently, agencies and brands have been vocal related to higher viewability standards. However, the issue has been hyped largely due to misinformation. By reading this POV, you will learn about the leading third-party measurement partners in the landscape, how viewability is measured and our recommendations for clients.

MEASURING & REPORTING

ON VIEWABILITY

We engage in ongoing conversations with both IAS and MOAT in the spirit of collaboration. Based on the way their technology works, these third-party measurement players cannot always measure the viewability of an ad. When data cannot be measured, IAS and MOAT will often report that as “not seen” or “unmeasurable” impressions which can skew the numbers.

Measurement

Viewability is measured using several different techniques (pixels, MRAID, SafeFrame and
TopFrame). Most providers used a tiered tag that tries multiple methods to collect the data.

Scoring

Viewability scores are done pre-bid and post-bid. Let’s dig into both.

Pre-Bid

Our platform uses IAS’ Pre-Bid Integration to utilize URL level scoring. Our clients receive reports based on pre-bid scoring for display, mobile and social with the exception of video and native. IAS’ pre-bid scoring samples domains in order to predict an overall estimation of ad viewability.

It is important to note that IAS pre-bid scoring is reactionary data that is used to make an
assumption at bid time, and is the deepest level of integration that IAS offers. Based on this
scoring, our platform sets viewability campaign parameters at “good” (45% viewable), “better” (55% viewable), or “best” (65% viewable).

Post-Bid

IAS and MOAT use Live Measure Scoring for post-bid. From here, we can measure the real
viewability of an ad. However, it is important to note that unmeasurable impressions are not taken into account by IAS and MOAT when giving an overall score to campaigns which can skew reporting significantly. Impressions are unmeasurable when the viewability technology is unable to properly operate in order to rate an impression.

VIEWABILITY: EXPLAINED

SIMILARITIES & DIFFERENCES

A deeper look at IAS and MOAT

When data cannot be measured, IAS and MOAT will often report those impressions as “not
seen” or “unmeasurable” which skews the results. Additionally, their reports include
exchange domains – impressions attributed to domains that reflect the exchange(s) they came
from. This results in improper attribution when it comes to viewability. For example, we see
“audienceGPS.com” as a domain, which shows that the tech is unable to properly break out the level of detail needed to accurately report.

So what explains these challenges? Third-party tracking tags are generally placed in what is
called a “hostile” or “unfriendly” iFrame. As defined by the IAS “a hostile iFrame refers to an
iFrame that has a different domain from the publisher page. These iFrames do not allow
JavaScript inside the iFrame to get information about the publisher page.”

This matters because an iFrame that has a different domain from the Publisher page will not
properly pass information back to these tag placements. In real time bidding (RTB), there are
multiple levels of iFrames each with their own tag and domain. And remember, hostile
iFrames do not allow JavaScript inside the iFrame to get information about the publisher
page. This is why third-party measurement players cannot always measure the viewability of
an ad, especially ads served through RTB.

VIEWABILITY: EXPLAINED

Their strength lies in measuring viewability of an ad in the browser. However, they use a Flash-based pixel which has a diminishing user base and does not perform well on mobile. The amount of people using Flash on mobile is not a solid representative sample to base post-bid viewability. Interestingly, IAS plans to go black-box with their data because pre-and-post-bid data sets are not lining up.

Key considerations include:

• Use of a Flash-based pixel for large portion of data collection along with mobile SDK
(the mobile app version of a pixel tag)
• Some in-app views hardcoded 100% viewable
• Struggle with measurability
• Reports include exchange domains

Their strength lies in measuring viewability of an ad in a mobile app.

Key considerations include:

• Use of a GIF-based pixel along with mobile SDK
• Most in-app views show as “unmeasurable”
• In-app must use their SDK
• Reports include exchange domains
• Claim significantly higher measurability

Both partners have their strengths and limitations to consider.

COMMON MISCONCEPTIONS

We’ve compiled 3 common myths to provide you the answers.

MYTH: IAS’ pre-bid integration judges viewability at bid time.

Actually, it’s a file of average scores used to make an assumption of how viewability
will be rated pre-bid.

MYTH: Mobile measurability can be guaranteed at 95%.

Often third-party measurement companies claim to guarantee mobile measurability
at a high rate. Measurability changes quite a bit depending on what type of traffic
you are looking at. For example, with MOAT as well as IAS, we have seen that most
in-app mobile traffic impressions are reported as unmeasurable. This means that
the viewability technology was unable to properly operate in order to give a rating to
the impression. This argument is quickly negated because at Simpli.fi, we do not get
charged on in-app unless the ad was viewable.

MYTH: These companies can now pass data between iFrames.

Viewability tag data does not get passed easily through iFrames which impacts the
ability of a third-party measurement company to track viewability. Often, the layers
of tracking can interfere with reporting – which causes the ad to return as
“not viewable.”

We have all heard of the industry’s quest for 100% viewability. Even for the most future-thinking, a guarantee of 100% is not achievable within publisher’s current revenue models. Their supply of ads would have to increase substantially to keep their current revenue
stream so as you can imagine, publishers are in no hurry to guarantee these numbers.

Rather than focus solely on viewability, we recommend clients consider the overall performance of a campaign and other KPI’s related to the engagement and effectiveness of an ad. Before launching a campaign, we discuss measurability with clients and treat measurable impressions as a sample set for the whole.

We support IAS and MOAT, but it’s important to understand both the
strong suites and limitations of each. We have data of our own to
help uncover the fuller picture.