Is Long Tail Assumption Valid?

[editor]

A recent study by Anita Elberse from Harvard Business school disputes the Long Tail Assumption, made popular by Wired's editor Chris Anderson. The Harvard study (tinyurl.com/3rg5gp) shows that in the on-line music and movie business, blockbusters are still very dominant.

What do you think?

[TimManns]

my two pennies worth...

I'm not in retail, but I think the choice of product or services (the breath *available*) always plays an important role even if they are rarely used. In my company (telco) we offer a huge range of services, and far more commonly see the 20/80 rule (often stronger) in quantity and breath.

But I believe that if we 'slim-lined' our business and offered only the most common and used services or products we would consequently suffer. Even if a customer only uses a 'long-tail' or rare service or buys a product once, if they can get that product elsewhere they will and probably stay here. If you can't service the customer's *every* need they will most likely always purchase from someone that can.

I had a quick read of the article. Maybe I'm missing the point, but it seemed to focus on 'sales events' rather than on-going customer purchase behaviour. I'd like to know how many customers that buy some of the most common product *never* buy some of the rarest....

Cheers

Tim

[editor]

There is always choice, but the premise of the Long Tail idea was that in the on-line world there is almost infinite choice, and more and more people would choose niche products. The Harvard study says that blockbusters are still very dominant, and that although the tail is Long, it is very flat - very little demand. Another interesting observation: people who get items from the tail are those with larger consumption. Occasional users tend to stick to most popular products, regardless of what is available in the tail.

However, this study did not mention if recommendations were given to users, like in Netflix case.
I wonder how they would skew the results.

[Eric Siegel]

I think the question is, given product recommendations that eventually gain user confidence, can there be a cultural shift away from purchasing by trend. This is not a question that can be answered with statistics, unless in the positive (over data where recommendations were deployed successfully) - it is more in the realm of psychology and other fields.