“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by Jim Parkinson, chief digital officer at Valassis.
I don’t come from this media world, so I learn something almost every day.
What I’ve recently found very interesting is the intersection between computer science and media delivery. There is this fairly complex and interesting relationship among data, technology, media and analytics that we all optimize to gain advantage. These interactions all change at a fairly high rate of frequency and create opportunities to succeed and fail fairly regularly. The other interesting function to this equation is the way clients buy digital advertising. This is truly a market where opportunity is the mother of invention, not necessity.
Take location-based advertising – this year’s phrase that pays. While many people play in this space, the most important driver is the technology that we all consume every day, including our cars, cell phones, computers, tablets and games. A pattern forms fairly quickly when you start mapping people’s physical traffic. In order to drive advantage, you have to focus not so much on the pattern but what the pattern tells us about you and the people around you. As we watch these patterns and compare them to patterns of others, we start to predict how you might proceed and what you might care about given where you are.
A Moore’s Law For Advertising?
How fast is that? One gigahertz is equal to a 1 billion cycles per second. A clock tick is a measure of work by a CPU, which is called a cycle. Does this mean we should have a Moore’s Law for digital advertising? I doubt that is possible given the intersections we have described, and measurement is purely subjective based on the desired actions at this point. However, David McMullen did a fairly nice job describing the problem in his March 2012 article, “Moore’s Law Meets The CMO.”
We see the clear intersection of data (where you are, where you have been), media (are you on a phone or tablet?), analytics (what will you care about?) and technology (make a decision, get the media, get the creative). It sounds simple until you try to successfully do it billions of times an hour.
This is why digital advertising is so interesting to computer scientists: It is the ultimate lab to figure out how to teach a machine to learn, and then act upon that learning in near real time.
When Failure Is OK
The acceptance of failure rates opens the door for another technology advancement. No other vein of computer science could embrace such a high failure rate. This opens the door to solving massive computer science problems. Big data would be nowhere without advertising. I would argue that most of the file system and database innovations are a direct result of these and other advertising media intersections. We will continue to enhance our near real-time analytics and machine decisions as the opportunity grows. Machine learning is coming alive again as a discipline in engineering circles.
Connection and memory technology will need to continue to evolve as the problems grow and geographic distribution continues to expand.
Tackling The Human Dimension
While the laws of math and physics, ethics and the opportunity to understand humans will always challenge us, behavior and the impacts on the behavior grow exponentially. Really, when we understand this intersection we are really trying to understand and predict: “What do you care about, based on where you are, and what will modify your behavior?”
There is no limit to what we can learn and what problems we will solve for brands that want to use digital media. This intersection will drive advertising and technology at a rate few other markets have seen. It forces engineers to build machines and software that are rivaled only by the human brain in their complexity and speed at which they must act.
The limiting factor will be good ethics, rather than science. Just because we can, should we?
That is a question for another day.
Follow Jim Parkinson (@jimparkinson), Valassis (@ValassisVCI) and AdExchanger (@adexchanger) on Twitter.
