I received an email from Amazon this week touting women’s shoes by Marc Jacobs. The subject line of the email was: “Marc by Marc Jacobs: Have the New Spring Collection Tomorrow with Overnight Shipping” This would have been great to get free shipping, with the exception of the fact that I don’t wear pumps, or “Ballerina Flats.”
Hold your wise cracks please. ;-)
I often think of Amazon’s email as one of the more targeted, accurate and restrained program out there. Since I buy marketing books, they often offer me more marketing books. That’s a great example of using my purchase behavior and their collaborative filtering technology to present me with offers they have good reason to think I might buy. It works too. Amazon has probably received more of my repurchase dollars than any other retailer.
They have sophistication to their engine, clearly. For example, I buy a strange array of music from Pat Metheny to Radiohead to Phish, yet they seem to be able to discern my music tastes, or at least their offers are always appealing. They either sell me deeper catalog for a certain artist, or leverage the “customers that bought Pat Metheny Group “Still Life Talking” also bought Wes Montgomery. Both are effective because they appeal to my stated interests (in that I’ve acted upon them and they observed these preferences in my purchase activity).
In a silo, however, behavioral or purchase analysis can go wrong if it isn’t married with other data. I’ve bought gifts at Amazon for my wife before, but it certainly would have helped them to know (or pay attention to) the fact that I’m male when they sent me the spring shoes offer. Maybe it was an error, but I actually may have clicked if instead of seeing an array of open-toed and thong sandals I was presented with a loafer, tennis shoe or golf shoe.
Hold your wise cracks please. ;-)
I often think of Amazon’s email as one of the more targeted, accurate and restrained program out there. Since I buy marketing books, they often offer me more marketing books. That’s a great example of using my purchase behavior and their collaborative filtering technology to present me with offers they have good reason to think I might buy. It works too. Amazon has probably received more of my repurchase dollars than any other retailer.
They have sophistication to their engine, clearly. For example, I buy a strange array of music from Pat Metheny to Radiohead to Phish, yet they seem to be able to discern my music tastes, or at least their offers are always appealing. They either sell me deeper catalog for a certain artist, or leverage the “customers that bought Pat Metheny Group “Still Life Talking” also bought Wes Montgomery. Both are effective because they appeal to my stated interests (in that I’ve acted upon them and they observed these preferences in my purchase activity).
In a silo, however, behavioral or purchase analysis can go wrong if it isn’t married with other data. I’ve bought gifts at Amazon for my wife before, but it certainly would have helped them to know (or pay attention to) the fact that I’m male when they sent me the spring shoes offer. Maybe it was an error, but I actually may have clicked if instead of seeing an array of open-toed and thong sandals I was presented with a loafer, tennis shoe or golf shoe.
ExactTarget Feed
Powered by Compendium Blogware
Comments for Behavioral Email Targeting Can't Exist in a Silo