Viewability: The Evolution of Digital Advertising: From Direct Response to Branding

Zenith Optimedia reports that spending for online advertising (29% of all Australian ad expenditures) is now almost equal to the amount of money spent on TV ads (31% of all ad spending). However, 74% of that digital spending represents search and related direct response tactics.

When it comes to branding, advertising on the Internet is equivalent to only 24% of Australian TV spending.

One of the reasons for this imbalance is that the Internet has its roots in search. But another perhaps more important reason is that we’ve used the wrong metrics to evaluate ad effectiveness. Take the click on a display ad. Going back to the beginning of the digital world, it was believed that a click would reveal the effectiveness of an ad. Yes, that still works well for search ads (where click rates remain material) and might have been a good approach back when click rates on display ads were running at levels of 3% to 5%, but those days are long gone. Today, Doubleclick reports that click rates for display ad campaigns average only 0.1%, which means that only one in a thousand ads in a campaign are clicked. Does this mean that digital ads have no value? The answer is a resounding “no”!

The problem isn’t a lack of effectiveness; the problem is the methodology being used to evaluate their effectiveness. A branding ad doesn’t have to elicit an immediate reaction to be effective. Witness the effectiveness of TV branding ads. What’s needed is a measure of the latent effect of advertising. That is, the cumulative impact over time that comes from frequency of ad exposure.

Back in 2009, Comscore conducted a seminal series of studies* showing that despite low click rates, display campaigns did indeed work. We did this by comparing the behavior over time of people exposed to brands’ display ads with the behavior of a control group of people who were not exposed to the campaigns. These studies showed that display ads were indeed able to lift site visitation, the number of trademark search queries, and both online and offline sales – all despite very low click rates. Welcome to the new world of digital brand advertising.

Some two years ago, another important step on the digital branding path was taken when the three U.S. trade bodies that represent the digital advertising ecosystem: the IAB (publishers), the ANA (advertisers) and the 4As (agencies) announced an initiative named “Making Measurement Make Sense” (3MS). The focus of 3MS was on establishing a new definition of a digital ad impression, which would only be counted if it was “in-view” to the consumer. There are several reasons why an online ad might not be in-view, the main one being that the ad is loaded on the page by the ad server but below the screen and the consumer doesn’t scroll down far enough to see the ad before clicking off the page. However, another cause is non-human bot traffic (NHT) that fraudulently captures ad impressions. The 3MS objective was to eliminate such low quality ad inventory and make digital directly comparable to TV, where “opportunity for the consumer to view” is an accepted tenet of brand advertising. So, today, a metric has emerged that holds great promise for the future of the Internet advertising industry. It’s called “viewability”.

Here at Comscore, we’ve measured the viewability of billions of digital display ads from campaigns around the world. The results were startling: on average, only 46% of ads were ever in view to the consumer. Obviously, this is a problem because an ad that’s not in view has a zero chance of affecting consumer behavior. Interestingly, the viewability rate was higher on premium publisher sites (53%) than on the ad networks and the exchanges (31%), with a substantial part of the difference being traceable to higher NHT on the exchanges.

There’s substantial upside to be realized if viewability can be improved. Comscore research has shown that increasing in-view rates can generate out-sized increases in ad impact. For example, Kellogg’s has realized a 75% increase in sales lift by increasing its viewability rates by 40%. The trick is to monitor how digital campaigns are being delivered in real time on a publisher-by-publisher basis so that corrections can be made as necessary while the campaign is still running.

Using tools and reporting systems such as validated Campaign Essentials (vCE) from Comscore, not only can viewability be monitored in real-time, but also exposure frequency and the accuracy of demographic targeting. Once the campaign delivery is being measured and results shared with individual publishers, ad money can be shifted to the publishers and placements that can deliver the media plan as intended. The benefits of doing this are substantial. Kellogg’s has reported that the ROI from their investment in digital advertising has increased by factors of five times to six times since they and their agencies have been optimizing digital campaigns in-flight.

As one can imagine, the use of viewability metrics has generated some advertiser demand for guaranteed digital audiences – such as those they are used to receiving in TV buys. While the jury is still out as to how Australian publishers will respond, it’s certainly clear that viewability is an issue whose time has come and one which promises success in attracting more branding advertising to the Internet.