Who could have predicted the pet rock or the hula hoop or Google or Facebook or the rebound and dominance of Apple?
What can data-mining and behavior-based marketing really do, where does it work most effectively, and when should marketers drop it and just follow their instincts, or in other cases, follow a counterintuitive plan to stick with their base business model?
The value placed on research data will continue to grow and the need for people with math skills will grow with it. Simple sales data has always been critical to demand planning. Marketers are always pressed to find a way to get an edge on the competition. Data creates business insights just as a physical examination gives a doctor insight into a patient’s condition and best treatment.
All Data is Historical.
I heard a Polaroid sales manager say one time, “History is like examining a dead dog’s corpse.” His view of business was entirely based on inventory push. The more you sell in, the more you sell through. The year before Polaroid’s biggest single-year decline in business was the year after Polaroid had the world’s leading selling camera.
Were forecasting failures the cause of Polaroid’s huge decline and eventual bankruptcy filing? In some ways, preventing Polaroid’s financial decline was like preventing the decline in the sales of the horse-drawn-buggies during the growth in the sales of cars. Polaroid was deeply invested in the manufacturing of the instant-movie systems when the camcorder came into the marketplace. For Polaroid even to transition to a digital system was complicated by the fact that its foundation in capturing images was not in electronics but in optical physics (light properties, dynamics, transfer, and relationship to other properties of physics).
BIG DATA, SHOPPER DATA, PANEL DATA
The term “Big Data” refers to Internet data. Internet data mining is really hot right now. Although the Internet makes anything it carries in an open environment appear to be current, by the time any forecaster or forecasting system accesses that data, the information is historical and only as valid as the consistency between when things happened and when they are going to happen. Shopper data is, of course, anything bought and tracked to the consumer. Panel data, in its simplest form, is based on surveys of the same people (a panel) over a period of time. I defer to the experts to fine tune my definitions.
What has changed in the data arena is not just the idea that people can predict what will happen, but that people can use data to alter what will happen in the future. Behavior-based marketing finds its limitations when marketers are asking the wrong questions. Counterintuitive marketing seeks to find the right questions. Jamie Turner blogs, “Consumers aren’t in need of 1/4 inch drill bits. They’re in need of 1/4 inch holes.”
Polaroid’s attempt at selling the instant movie camera system is perhaps an interesting case in point. Answering the question, why can’t I see my pictures right now, Polaroid invented the instant camera. Answering the question, why can’t I see my home movies right now, Polaroid rolled out a camera, which required its own player. The question that consumers were apparently asking was why can’t I watch home movies on my television?
Who could have predicted the pet rock or the hula hoop or Google or Facebook or the rebound and dominance of Apple? No one, I suppose. Yet I am sure that researchers are busy working on those and other marketing and business questions right now.