Viewability is a Poor Proxy for Quality

On May 17, 2016 in London, Operative hosted a publisher breakfast at the beautiful Charlotte Street Hotel with the theme, The Future of Measurement: Viewability, Time Spent, and What Comes Next. The event was standing room only with over 30 publishers in attendance, including The Economist, Financial Times, Bauer, and Schibsted and was sponsored by Moat and SpotX.

The event got kicked off with a welcome from senior leadership at Operative Moat and SpotX. However, things really got lively once publishers came on stage for a discussion panel about the viewability and the various metrics that can best determine campaign and ad quality. The panel featured Alistair Smith from the Financial Times, Dom Perkins from Time, and Ashwin Srindhar from The Economist, and was hosted by Ciaran O’Kane from ExchangeWire.

Both the Financial Times and The Economist have made headlines recently around their industry leading approach of selling on Time Spent and Attention based metrics, as an antidote to the constant dissension around how viewable impressions are defined. The initial discussion focused around how “Viewability” has become a watchword for “quality control” – across the 7 trillion impressions that are available to buy (“Let’s just call it infinity”, quipped Perkins). This almost endless supply of inventory, caused by the typical approach of just throwing more and more ads on a page in the interests of revenue growth, has a major impact on the way publishers are able to maximize inventory revenue.

There was general agreement, however, that the official definition of viewability for desktop was not making much of a dent in that huge 7 trillion volume of supply. The Current definition only knocks out maybe 20-30% of that 7 trillion and only provides 1 second of visibility. This doesn’t necessarily support the increased quality claim, and doesn’t justify the price premium that ought to be charged for truly viewable impressions.

Both the Economist and Financial Times have addressed this by focusing on time spent and attention. They will sell a package of hours to advertisers, where the consumer has seen an ad that has been in view for 5+ seconds. This can be proved to be more effective and provide a greater ROI, based on the repeat buys that both publishers have seen. However, a challenge that comes up with this approach is that buyers “can’t fit it into their spreadsheet”. To get across that barrier, technology companies on both sides of the transaction need to provide flexibility. Then they need to support new processes and workflows such as optimizing and capping based on time, rather than the number of ads served. Only then can Operations support this type of selling in a scalable manner, a crucial need for a time-based campaign, where they have to handle this across multiple publications.

Other insights that were shared:

• When anyone is asked about whether or not 100% viewability is possible the answer that really there is no such thing. Only 4% of UK ads are currently 100% viewable.
• Digital should not be lumped into a single bucket; Mobile and Desktop should be considered different mediums at the best of times, and viewability just accentuates that difference
• By using deterministic data some publishers are actually taking ads off the page, to make the ones that are left behind more effective

The panel concluded with the acknowledgement that we still have a long way to go in this discussion, and that it needs to be a joint process where publishers work to provide ways to capture the attention of the consumer that the buyer can use to put something effective in front of that consumer, while buyers can be open to ways of transacting that are not laser focused on what fits into a spreadsheet and into variations on the traditional impression model.

Additional Resources:

• Operative/Moat Integration Collateral
• Future of Programmatic Whitepaper
• Moat Website
• SpotX Website