GroupM: The Elephant in the Room: Why Advocating Attribution Modeling Can Strain Credibility - Chris Copeland - MediaBizBloggers

By GroupM InSites Archives
Cover image for  article: GroupM: The Elephant in the Room: Why Advocating Attribution Modeling Can Strain Credibility - Chris Copeland - MediaBizBloggers

There is no topic more polarizing in the digital space than that of last click attribution models. To be against reform means you are either a bleeding-heart search guy who simply wants all the credit or a Neanderthal, although some might ask what the difference between the two is exactly. The occasional wise guy will point out that a true attribution model would include off-line impact, but there's always a spare stapler or desk chair to use to silence that kind of contrarian view to the real point.

The real point is about two truths which, contrary to popular belief, are not advertising creations: money and power. Search marketing represents more than 50 percent of all digital spending. That kind of money is something many in the display sector are quite interested in controlling. It is interesting to note that in general, the largest buyers of search are third party bid management companies and direct advertisers, not agencies. Agencies, in general, spend more on display than on search because of who their clients are. Those clients are more traditional (read not direct response) advertisers that want to create awareness and impact brand perception.

And that brings us to a credibility challenge that these agencies, especially those who engage in traditional forms of media planning and buying, face when they thrust forward the need for better modeling. The starting rationale for attribution is to give credit to the right person. In basketball, an assist goes to the person who distributes the ball immediately prior to a field goal. Hockey gives credit to the two people who were involved prior to a goal being scored. In tracking statistics, they do not report whether someone was the first or second pass prior to the goal. Unlike sports, media seems intent on trying to solve the challenge of settling on the right attribution model to give the proper amount of credit to each participant. Are they all equal? Does the final act deserve something more? Does it really matter?

Let's take the case of a professional basketball player to explore why knowing who was involved may be more important than worrying about the level of credit each participant deserves. Throughout the history of the NBA there have been few point guards who were better at handing out assists than Jason Kidd. Kidd, a 16-year NBA veteran, ranks sixth all-time in assists per game with 9.2 per game. By the measure of assists alone we can assume that Kidd is one of the best of all time. Unfortunately, he also has the distinction of being one of the worst shooters of any elite player in the history of the game. His 40 percent field goal percentage is especially poor when you compare him to every other elite point guard on the list. Kidd ranks sixth on a list where the average FG percentage of the guys ranked above him is 48 percent. That's a huge variance. Translation – you want Jason Kidd passing you the ball, not shooting it. The problem is most other elite players at his position could convert if needed with much greater certainty than Kidd. So, he's truly display to search in the assist and conversion discussion. You want him in the mix and close to the end outcome, but not as option one to finish.

It is virtually impossible to consistently know how much intention a display ad established that led to a click. In absence of stronger panel or mind reading data, we just don't know. So, it would appear that when someone advocates a model it is usually a means to justify a reallocation of money and power. The problem, as I see it, is not the need to give credit but the digital scrum taking place at the expense of better performance. It is well documented that search is a great fulfillment vehicle. Search plus display buying, and now, through research we conducted last year, search and social combined have proven that when paired together the outcome is better than on its own.

So, why are we playing a game of either/or with digital channels? I think it is fantastic to understand how a specific media placement contributed to the end outcome, what link in the chain it was, and, if possible, associate repeat actions with future buying decisions a planner may make. My problem is why are we jeopardizing any money or power in digital at the expense of other digital elements. A new study from Fleishman-Hillard and Harris Interactive discussed the continuing disconnect between digital spend and digital influence on consumers in comparison to other media. I have more thoughts on the validity of that thinking to come, but for now let's focus on the impact of search as fulfillment in relation to traditional media.

If we accept that search is fulfillment for display and therefore warrants funding to score the conversion, then what is stopping us from doing the same thing with traditional? Shouldn't we be modeling the likelihood of impact from television, print and all other traditional media that drives to search at the end of the day? If yes, then our funding models when we think about attribution should never be about taking away from search to fund display. Rather, the model should ensure that we have the right investment in display for the influence it has on search, as well as enough search share to fulfill what is driven from off-line channels.

There is no perfect model for attribution in a digital-only environment. Even worse, the debate pits growth in digital against itself while ignoring the importance of digital to extend the investments being made in TV and other traditional channels. No one in their right mind, even a die-hard search guy, would suggest that search is the be all. But if it is the end all for many online conversions, then it is vital that the funding models evolve to include all areas of influence and not just stay bound to an online silo.

Chris Copeland is CEO of GroupM Search for the Americas. Chris can be reached at chris.copeland@groupm. You can also follow Chris on Twitter @SearchBoss.

Read all Chris' MediaBizBloggers commentaries at Musings from GroupM - MediaBizBloggers.

Check us out on Facebook at MediaBizBloggers.com
Follow our Twitter updates @MediaBizBlogger

Copyright ©2024 MediaVillage, Inc. All rights reserved. By using this site you agree to the Terms of Use and Privacy Policy.