Viewable Yet Unseen: It’s Time For New Engagement Metrics

philschraederupdatedData-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.

Today’s column is written by Phil Schraeder, chief operating officer and chief financial officer at GumGum.

Being viewable and being seen are not the same.

Yet today’s engagement measurement metrics do not take this key difference into account, to the severe detriment of brands that only discern value from their media investments when consumers actually see their ads.

It’s time to look beyond viewability and create a more sophisticated set of engagement metrics. Viewability is only good for potential. Actually seeing an ad, on the other hand, is a much more active process.

Nearly two years after comScore first called out the industry for wasting more than half of advertisers’ investment in display, an army of analytics vendors now stands ready with varying flavors of measurement and accountability.

The Media Rating Council (MRC) has established a viewable impression standard on which we can all agree, calling for desktop display ads to be considered viewable if 50% of their pixels are in view for at least one second.

But while 50% of an ad in view for one second is a fine technical standard, it hardly constitutes a brand metric. It’s not very realistic to say that someone who sees a little more than half the pixels of an ad for a little more than a second has really processed the image, much less the message behind it. Marketers need to focus on what takes an ad from viewable to being seen.

Uncertain Future

Unfortunately, the sheer number of measurement vendors – and the wide discrepancies between them – has left the industry still grappling with how to transact on viewable impressions. This led the IAB to recently publish a “State of Viewability Transaction” report [PDF], which said an acceptable viewable impression threshold is 70% of ad pixels in view. The IAB also said that if a campaign doesn’t achieve this threshold for measured impressions, publishers should “make good with additional viewable impressions until the threshold is met,” placing pressure on publishers to ensure a certain degree of ad performance.

Whether directly sold or delivered programmatically, advertisers deserve to confidently invest in viewable impressions. Transacting on viewable impressions will be figured out between buyers and sellers, their respective measurement vendors and the industry’s trade bodies. In the meantime, we should focus on the effectiveness of those impressions, not the least of which is making sure the ads are actually seen, not just technically viewable.

Better Metrics

When everything becomes viewable, other metrics that go beyond the concept of viewablity are needed.

These kinds of metrics include Millward Brown’s ability to measure post-impression consumer search and site visitation behavior, which can track whether a customer is searching or researching a product related to a specific ad and measure the ad’s influence. There is also Nielsen Online Brand Effect’s measurement of brand lift, as studied through exposed and unexposed panel methods to determine if ads actually led to conversions.

Other useful tools are engagement metrics from MOAT, such as hover rates and time in view, which can offer valuable insights into how customers specifically interact with content. This can tell marketers how customers are browsing and how different kinds of ads can influence that browsing experience.

With the information that these metrics provide, brands can measure the lasting impact of their ad campaigns, in terms of brand awareness and customer behavior. This tells them which ads are having an effect on customers in ways that are beneficial to the brand, not just the advertiser or publisher.

These are the metrics brands should care about most. Everything on a page is viewable, but being viewable is not the same as being seen. Anything and everything in front of a person is viewable, but it’s the elements that are engaging, interesting and eye-catching that are actually seen.

The majority of an ad’s value comes from its ability to draw the viewer in. In an age where consumers are bombarded with more content than they can process, content that is seen by consumers captures and holds users’ attention long enough to process the message is what truly defines success.

Follow GumGum (@GumGum) and AdExchanger (@adexchanger) on Twitter.

Enjoying this content?

Sign up to be an AdExchanger Member today and get unlimited access to articles like this, plus proprietary data and research, conference discounts, on-demand access to event content, and more!

Join Today!


  1. While I’d agree that being ‘viewable’ does not constitute brand engagement I think there needs to be caution in suggesting that they are part and parcel of the same thing. Viewability is merely a metric that tells advertisers whether their ad has/had any chance of being seen, and in turn is a flag as to whether they should continue buying from that source/domain, and it is [becoming] widely accepted that the seller should take responsibility for an ad if it was not viewable, i.e. make good with viewable impressions or not charge. But assuming an ad impression is deemed viewable the engagement that takes place after that is not necessarily the problem of the supplier. It is, of course, prudent for an advertiser to monitor brand engagement and to optimise a campaign towards it, but we don’t want to get to position where brands begin to believe they should choose not pay for an ad impression if the user didn’t engage with it.

  2. This article brings up the important points that “in-view” metrics aren’t the best way to measure the success of a campaign, and the alternative metrics suggested are a good thing. However, this story does not get at the real viewability problem – that the ad tech ecosystem and publishers won’t change unless brands change the way they pay their partners. Vendors and publishers have monetary incentives to hide real performance from marketers because they get paid on volume, not performance.
    As long as marketers continue to allow media teams to pay CPM prices, then most of the ads placed will be misplaced. Once marketers focus on the metrics that actually drive performance and pay based on them, metrics like viewability will be informative rather than distracting.

  3. The problem with coming up with standards is, who is coming up with the standards? If you look at IAB, they are comprised of many different companies with many different ideas on solving ad fraud. Getting them all to agree to something that all of their technology will agree with, means coming up with the lowest common denominator; which is not very useful.

    Like you said, seeing 51% of a banner for 1 second is the standard. Think of it this way, it takes 250ms for the human eye to focus on an object. That gives you 750ms to process the image, which is barely enough time. However, if it’s only half of an image, it takes even longer to figure out what the image really is, and 3/4 of a second is not nearly long enough!

    With this standard, I have seen fraudsters, load up a page of banners, spoof their referrer, and rotate the page every few seconds. They then send a view-ability bot to that page and rotate it every so many seconds. So it meets the view-ability standards, but still does not help the advertiser.

    I have been preaching for years for my advertisers to focus on their backend metrics and help them to understand the best metrics to focus on. You have to pay attention to different metrics and continually evolve on what metrics you focus on.

  4. Quite simply viewability is essential table stakes. If the ad can’t be seen the targetability, content relevance, responsiveness, engagnement etc. and all such metrics are merely theoretical. And, importantly, these challenged viewability metrics cited are for computers; the mobile viewability world is even wobblier. Persuading advertisers to transfer vast budgets from TV to digital is hampered by this lack of OTS verification.