Recommendations for the Person, Not the Purchase

Last Saturday night, I was in the mood for a movie; I didn’t have anything particular in mind, so I hoped that Netflix would have something merchandised or recommended. I wanted to go to their site and think, “Yes, I’ve been dying to watch that!” (This is a sign of great merchandising, and I often feel that way in the B&N store and on several ecommerce sites.)

Netflix had very little to say – well, to accurately say – about what I should watch. Granted, all they had to go on was my rental of Friday Night Lights, a movie about lesbian acrobats, and the ratings I’d added when I joined (some of movies I’d watched a decade ago). I thought, “I should just call Carrie and see what she’s watched and liked recently.” Why? Because Carrie always knows what I will love: we are both part of the same generation, have grad degrees, read the Sunday Times, live in an urban area, and read voraciously. Her recommendations aren’t based on what I’ve bought in the past, but an understanding of my persona.

Persona recommendations are so much more useful than last purchase recommendations. Haven’t most of us bought something for a friend’s baby on Amazon, only to get bombarded with stroller recommendations? On BN.com, recommendations are based on last purchase, so if you last read The Hunger Games, you’re sure to love Vampire Academy – even if you’re a 30 year-old guy.

But what if purchase history was a much smaller part of a holistic formula? I look at my Goodreads feed, and so many of my friends’ books overlap with mine – friends that aren’t that different from Carrie and me. Yes, I like cooking murder mysteries more than most, and Amy tends to read more theory, but 80%(ish) of what we love is similar. What would happen to B&N’s recommendations if instead of saying, “She liked Hunger Games,” they said, “She is a professional, Gen-Xer, urbanite, etc., etc.”?

What I did like about Netflix is how they asked questions about what I liked, and yes, that’s a start. But they didn’t ask much about who I am – and who I am determines what I like, not the other way around. Ultimately, book recommendations (actually, any ecommerce recommendations) would be more powerful if they looked at an entire persona, not just last purchase.

Carrie, thank you for the Portlandia recommendation.