Today’s column is written by Sudha Reddy, director of ad operations, West Coast, at xAd.
While it seems that everyone is talking about header bidding, fraud, cross-screen, viewability and data, I see the biggest issues facing the industry being discrepancies and the need for standardized measurement of key variables.
Ask any ad ops person about his or her biggest career challenge and you will probably hear that discrepancies are the No.1 factor impacting their lives. Many have nightmares about waking up to solve a whole new boatload of discrepancy issues in their campaigns. To complicate the challenge, new companies constantly pop up to solve the same issue and create yet another new way of measurement.
For example, an advertiser’s agency might plan to launch an ad campaign using its preferred demand-side platform (DSP), which uses several exchanges to serve ads. The advertiser may use DoubleClick DART (Dynamic Advertising Reporting and Targeting) as a billing source but its agency prefers Medialets tags and the DSP has its own measurement mechanism. Most publishers on the exchange probably use DoubleClick for Publishers as their ad server.
The advertiser bills the agency/DSP using DART numbers but these might not map with the Medialets or DSP numbers. The DSP also bills the publisher using the publisher’s ad server numbers but those might be different from the DSP figures.
To make matters more complicated, the advertiser might use multiple agencies that may have their own trade desks or preferred DSPs to serve ads. In several scenarios, measurement trackers for brand safety, location and viewability are also involved and add to the discrepancies.
Then there is the impression: Every vendor probably uses the impression definition a little differently. Some refer to it to mean a served impression that is counted when the request is received and served by the ad server. Others take it to mean a shown impression that is counted when the impression is 50% rendered. And some might refer to it as a loaded impression, counted when the impression is fully loaded and viewed by the user.
In most scenarios, publishers and their ad servers are the closest to the chain and have some control over when the impression is fired. Since very few advertisers and publishers directly interact in today’s ecosystem to set up these parameters ahead of time, should there be a default set of standard definitions across board?
Facebook recently revealed how it wrongly calculated video views, which pretty much summarizes the problem. If a giant company like Facebook can make such mistakes, how about the zillion other smaller players in the industry?
We have had the Media Rating Council and the Interactive Advertising Bureau (IAB) standardize some of the terms in the ad industry and also push the industry to adopt HTML5, SafeFrames, MRAID (mobile rich media ad interface definitions) and VAST (video ad serving template), which to some extent helped. But even after so many iterations, the discrepancies are only growing, not declining.
With new technologies and vendors entering the market every day, there seems to be no dearth of new trackers that could be added to campaigns nor of the number of discrepancies we are dealing with.
What exactly is the solution? Can IAB dictate that all publishers use one standard set of definitions and have these certified? Can we have exchanges and publishers that will across the board follow the standard IAB measurements?
I am not sure if any of these will solve the issue nor if the discrepancies will ever completely go away – just like the nightmares that won’t stop for anyone who works in ad ops.