Have you heard the chatter about “Math Men” taking over from “Mad Men?” Don’t believe it. Marketing remains more of a craft than a science, says MediaCom’s Oliver Gertz, EMEA Managing Director, Interaction and Programmatic Lead.
Don’t get me wrong: there’s no question that data driven marketing has completely changed the game. What started with overnight TV ad ratings has morphed into the average Demand Side Platform (DSP) spewing up to 80 data points for each ad impression we buy. Online microtargeting allows us to test thousands of microsegments against hundreds of message variations in real time, and advanced attribution modeling gives us greater insight into each touchpoint along the purchase path. What this means is that the analytical and data skills that have always been key to effective communications planning and media optimization are more valuable than ever. Consumer insights is evolving into audience segmentation embedded in today’s media buying tools. Media planners are transforming into data planners. Econometricians are now called data scientists and apply their skills to much bigger data sets.
What it doesn’t mean is that all media planners and marketing managers will be replaced by mathematicians, programmers and data scientists. What’s most needed is still a deep understanding of consumer behavior – what they care about, how and when to engage them and how to get better and better at it. Yes, we do need more programmers and data scientists. But their role will be to support the planners, optimizers and clients in making sense of more data, rather than taking over marketing altogether.
Losing Jobs to Algorithms
Who cares? We’re all going to lose our jobs to machines anyway, right? That’s not going to happen any time soon, either. Search engine advertising, for example, seems to have a very simple bidding setup: one or two vendors, a few ad formats, brand-safe environments and algorithms that drive media buying. That’s it. What we’ve learned, however, is that twice as much effort is required to spend a euro in search than in traditional display buying. The ability to optimize to the most granular keyword level while refining bidding rules and customized messages actually forces us to spend much more time in setup and optimization. The same can be said about applying real-time data, which must also be properly analyzed and optimized.
Another example: as we move into social advertising and programmatic, the options and data volumes will require more people to handle the workload. A traditional media buy of 100,000 impressions is based on the evaluation of three to five offers, followed by one insertion order to buy the ad space. A programmatic media buy of 100,000 impressions is evaluating millions of bid requests and deciding on 100,000 individual bids, all executed in real time based on data and predefined algorithms. We need algorithms to manage such complexity, but we also need more human involvement to drive those tools. And how amazing is that? For the first time, consumer-centric, data-driven marketing allows us to cherry-pick an individual consumer profile and determine the value of each consumer. It also helps us establish the price we are willing to pay to reach that consumer right now, and then show a customized message based on all we know about him or her. Time? 120 milliseconds.
Math AND Madness
Such a scenario is truly “Content and Connections” at its best. Yes, data and technology are the engine, but no engine can run itself. Smart, intuitive, analytical humans will be as important if not more important than ever if we expect to surprise and delight consumers in an increasingly competitive, noisy world.