• by Karl Greenberg , Staff Writer, October 27, 2015

Acceleration and change. That’s the auto industry. Besides being thrown on its head by Apple and Google entering the autonomous car fray, and Tesla’s disrupting both the product and retail sides of the business, automakers are trying to navigate the protean world of digital marketing. It’s forcing marketers to filter data to evolve their tactics intelligently and efficiently. 

There will be no shortage of talk about big data at Mediapost’s inaugural automotive conference, Marketing Automotive, on November 9 in New York. With androids behind the steering wheel of the digital combine, data is filling the CMO’s office, 24/7, like chaff from a harvester.

How to deal with that? Max Knight, VP, global marketing science servicesat digital advertising technology company Turn, who will participate on a big data panel at our event, points out that pretty much everyone is doing automated media buys at this point. “It would be incredibly rare to see any maker, or model, not doing a significant amount of programmatic buying.” But, he adds, the trick is how to understand how your buy is working.  

For Knight, managing the data back-flow from sophisticated, automated digital buying requires looking at that data through the right set of lenses. And that point brings to mind countless conversations I’ve overheard and had, in which marketers mull whether vehicle sales should be part of the attribution model for measuring the efficacy of a digital media buy. Knight argues that, while you can go ahead and do that in a way you couldn’t in the past, it might not be in one's best interest. “The bigger question is whether it's useful to do it. Obviously, if you have a good attribution model it helps you spend more effectively.” But, he points out, people can get really caught up in the idea of "I have to have a perfect attribution model. Everything has to be taken into account.” 

Knight says that attribution perfectionism is crippling. “You need a model that is good enough to help guide your spend such that you get more out of it. There is no such thing as a perfect attribution model because we can’t get inside people’s heads and say ‘What impact did this one experience have on you right now?’”

Paul Miser, digital lead at Lincoln Motors’ AOR, New York-based Hudson Rouge — who will also appear at the conference, and will present on the digital program for the new brand campaign for Lincoln featuring Matthew McConaughey — says getting that understanding requires elasticity: a campaign has to be able to stretch horizontally, well beyond the official window of the campaign. Digital-channel buys made with the help of automation technology ease all of that because it’s cheap and marketers can make changes based on impact, and fire it into the digital deep blue for a lot less money than a TV campaign. “It lets us manage the storyline.”  

An issue for the auto purchase cycle is a boon for data: cars aren't cans of beer. Shopping for cars is a long journey, and a unique one. You don't go to an assortment of sites to do product research on Lebatts Blue before buying a six pack. Consumers who shop autos are going to OEM sites, third-party sites, blogs, and social media to compare and research. Says Knight, "Very, very few people wake up one day and say ‘I’m going to go buy an F-150. I'm going to walk down to the lot right now, and I'm just going to buy it, and that's it.’” There's a long research period, and it is online. All of that adds up to data that consumers trail like confetti. 

And tracking granular data can wind up an exercise in counting pennies. Decent attribution doesn't require you to enumerate all the single banner and search clicks, and video ads and native experiences all the way to the dealer sales floor. Says Knight, “What they have to do is say ‘Am I driving the right types of engagement?’ They know at a top level that the more people they get interested in something, the more people they get to go look at a brochure, the more people they get to go find a dealer, the more people they get to schedule a test drive or something like that, then the more they're going to move.” Again, the data goal should be macro — understanding the consumer experience. After all, with all the changes going on in the industry, marketers have enough to worry about.