by Nigel Hollis
The promise of cost efficient, real-time targeting means that programmatic advertising has taken the digital ad industry by storm. Magna Global predicts the market for global programmatic will double between now and 2019. However, the evidence suggests that programmatic still has a long way to go before it delivers on its promise.
Kantar Millward Brown in partnership with comScore have recently published a report detailing the state of the art with regard to programmatic advertising and the conclusions are not pretty. Based on analysis of billions of impressions, from hundreds of campaigns monitored across the globe, we find programmatic buys today are falling short of their initial promise.
One of the biggest problems identified by the report is that viewability and invalid traffic undermine the effectiveness of indirect programmatic ad buys. This is a big problem because if an ad is not seen by a human being it cannot possibly work to build the advertised brand or drive sales. Particularly for video advertising comScore validated Campaign Essentials (vCE) found that impressions run through programmatic exchanges had lower viewability rates compared to those purchased through direct buys (54 percent versus the direct figure of 73 percent) and invalid traffic was higher (11 percent versus 5 percent).
Unfortunately things get worse when we look at the ability of programmatic buys to hit their designated target. Across countries more than half the ads fell outside their intended target group, with desktop in-target rates ranging from 38 percent for France to 50 percent for Brazil. Accuracy is further reduced the more specific the target audience becomes. Now 100 percent accuracy is never going to be possible – cookie data can be old and inaccurate, registration data incorrect and computers shared – but you might hope for better than 38 percent.
The end result of this inherent variability in viewability and accuracy means that the impact of programmatic campaigns is also highly variable as measured by our Brand Lift Insights studies. On average programmatic does drive up purchase intent better than all the campaigns measured – likely due to the ease of identifying people already “in-market” and also due to a greater focus on call-to-action messaging – but is less likely to drive key metrics like ad awareness and message association which are important for brand building campaigns.
I interpret this to mean that programmatic is better at hitting people already predisposed to buy a brand, but is less likely to create that predisposition in the first place. This said, there is huge variability with some successful programmatic campaigns delivering on their promise and others failing dramatically. One of the keys to success with programmatic is the quality and depth of data that is used for targeting. The better the data, the better the buy is likely to be.
It seems from this data that programmatic still has a long way to go before it fulfills its initial promise, but what do you think?