Has the Long Tail Been Cropped?
Lee Gomes of the Wall Street Journal has written an article where he does his best to debunk much of the “Long Tail” concept coined by Chris Anderson in his new best seller, “The Long Tail.” When I read the article, it all sounded like he found some big holes. But Chris has fired back with his own blog post on this subject and does an admirable job of defending his original conclusions.
I have not yet read the book, so I can’t give my own views on what Chris meant, but the article makes for good conversation. Case in point. Gomes challenges the 98 percent rule (where 98 percent of choices are chosen by someone) and gives examples where
Ecast told me that now, with a much bigger inventory than when Mr. Anderson spoke to them two years ago, the quarterly no-play rate has risen from 2% to 12%. March data for the 1.1 million songs of Rhapsody, another streamer, shows a 22% no-play rate; another 19% got just one or two plays.
I do have to admit I find the idea of a “rule” is hard to swallow and I can see wide variations from this 98 percent figure across different markets and vendors – and time periods.
Gomes also takes issue with the idea that the tail will at some point grow to equal or greater than the head.
I was thus a little surprised when Mr. Anderson told me that he didn’t have any examples of this actually occurring. At Netflix and Amazon, two of his biggest case studies, misses won’t outsell hits for at least another decade, he said. None of these qualifications are in the book.
Gomes continues with an example:
By Mr. Anderson’s calculation, 25% of Amazon’s sales are from its tail, as they involve books you can’t find at a traditional retailer. But using another analysis of those numbers — an analysis that Mr. Anderson argues isn’t meaningful — you can show that 2.7% of Amazon’s titles produce a whopping 75% of its revenues. Not quite as impressive.
Anderson answers this in his blog, referring to an earlier post that does explain quite well (at least as well as you can explain this kind of statistical example) the difference that comes when you look at the numbers from opposite sides. It is very easy to come to vastly different conclusions when you have lots of numbers to play with. You can read each and see who you agree with.
Another example Gomes tells us about seems to miss the point:
Another theme of the book is that “hits are starting to rule less.” But when I looked online, I was surprised to see what seemed like the opposite. Ecast says 10% of its songs account for roughly 90% of its streams; monthly data from Rhapsody showed the top 10% songs getting 86% of streams.
Here Gomes seems to be looking at the number of streams, not the number of customers or products. I’m sure you can take the exact same numbers and show that 98 percent (as it was two years ago, or 88% today according the WSJ article itself) of their tracks get played once a quarter. So now who is twisting statistics? Anderson says he told Gomes about Netflix data that says 95% of their 60,000 titles are rented at least once per quarter. This is the point of the Long Tail – that all the titles not stocked in a local store are being accessed through an online store that does have the massive inventory needed for this, not that 90% of their total rentals happen to be the top few thousand movies available anywhere.
I do have to say that we see this with the classifieds. We tested out a tool called HitTail and looked at the incoming search terms. The top ten terms accounted for less than 3,000 of 50,000 search visitors. The other interesting thing I got from this data was that the 10th most popular term had only 4x the traffic of the 100th most popular term, and 8x the 250th. There is a lot of demand out there for really odd and obscure things, and those who know how to tap that market will succeed online.
There is some demand for everything.
-
http://www.hittail.com Mike Levin
-
http://www.pipelinedeals.com JP Werlin
-
http://www.rabbitbites.com buns and chou chou (
