Readers may have been confused by last week's back-and-forth about the extent to which the Netflix DVD demand curve differed from its bricks-and-mortar counterparts (perhaps because the WSJ reporter confused some of the definitions). These charts may help make it clear. In each case "Industry" refers to Nielsen statistic for the industry at large, which is about 85% bricks-and-mortar retailers such as Blockbuster. "Netflix" refers to a data set from the online service in mid-2005.
This first one shows the difference in the demand curve for the top 100 DVDs. What you can see is that the Netflix curve is much flatter, which is to say that it's less hit-centric.
This next one shows the market comparisons for the top 10,000 DVDs. Today Netflix actually has more than 60,000 titles, but I wanted to show a proper head-to-head comparison so I've only taken it as far as the offline data goes (that's why the Netflix data doesn't add up to 100%). This is another way to show how much more evenly the Netflix demand is spread between hits and niches. Of course, the hits--top decile--are still the majority of the demand (hits aren't going away!), but at Netflix that top decile represents only 64% of the business, versus 90% for the DVD industry (mostly offline) as a whole.