Saturday, January 22, 2011

Stay Close to the Customer

A kiosk sponsored by Kraft Foods, incorporating Anonymous Video Analytics technology, and based on Intel chips. That item was arguably the goofiest item at this month’s 100th Annual National Retail Federation show in NYC. At least goofy at first glance. And it’s the first glance that counts in this case, since what the kiosk does is look at your face and based on what it sees, makes menu suggestions.
     Well, arguably I’d say the kiosk wasn’t all that goofy, since it did lots more, too. Punch in information about your meal times, swipe your grocery store card, and the kiosk goes beyond judging your gender and age to incorporate your purchase history before generating meal ideas. Want a sample of an item it suggests you buy? Ask with a poke of the finger on the screen. The facial recognition was really no more than a gimmick to catch lots of media attention.
     Still, if we take this Kraft kiosk concept seriously for a moment, it can serve as a reminder of the folly in depending excessively on technology to give advice to our customers or to give us advice about what our customers need and want. Stay close to your customer.
     Here are two more examples of technology that can get in the way of accuracy if you fail to exercise caution:
  • Brain scan technologies like Functional Magnetic Resonance Imaging (fMRI) and Steady-State Typography (SST) help us understand why people buy what we’re selling. But the brain scan technologies are imperfect. For instance, according to The New Yorker, a postdoctoral fellow in neurological sciences at University of California-Santa Barbara performed fMRI measurements as part of a brain study to show limitations of the methodology. The readings gave evidence of activity in a region of the brain associated with empathy. Almost surely a false reading, though. See, the brain was of a salmon. A dead salmon.
  • Sophisticated statistical procedures like Structural Equation Modeling can allow you to understand truths such as what motivates people to recommend your store to others. But because of the complexity of the techniques, you’ll be trusting the analyses to an outside consultant using a computer. When working with a consultant who is using SEM or if you or your consultant are turning SEM findings into real-world action steps, remember that you know your retailing operations better than your consultant or a researcher does.
Click below for more:
Interpret Brain Science Advice Cautiously
Call on Structural Equation Modeling
Look at Mean, Median, Mode, and Range

No comments:

Post a Comment