I promise not to prattle more about the long tail for a spell, but I indirectly had an interesting conversation with someone today on the subject. One of Chris Anderson’s examples — matter of fact, one that leads his original Wired article — is that of recommendation engines and they way they can push you out into the long tail.
You know, if you like “Into Thin Air” you’ll probably like “Touching the Void”. That sort of thing. (By the way, had a strange typo when I first typed the title of Joe Simpson’s great book, called it “Touching the Voip”. Would be a great presentation title for a harrowing wander through the halls of dead and dying VOIP companies. But I digress).
Anyway, the person with whom I was talking today argued that was all just data mining, which has improved a lot since the (and I’m quoting him here) “beer and diapers days”. (There is an old story that some stores incongruously group beer and diapers together in the store because they are both impulse buys.) The effect of recommendation engines is to group virtual beer and diapers together online: If you like this, then you’ll like that — and verily you wander into the valley of the long tail.
That’s true enough, as far as it goes. But is worth pointing out that the whole beer and diapers idea (which I admit to have wrong-headedly retailed once or thrice) is very likely just another urban legend:
Lynette Dyer (of Cogit) gave a very good presentation that played down much of the hype around data mining, emphasising the need for a clear business value proposition. She also debunked the ‘beer and nappies’ example, which she said was invented by Tom XXXX as a joke when he and others were doing some analysis for Wal-Mart.