There seems to be a bunch of problems with them making a definitive statement about internet business based on this research. First, the Nielson list doesn't i

There seems to be a bunch of problems with them making a definitive statement about internet business based on this research. First, the Nielson list doesn't include data from two of the G8: Japan and Canada. It also doesn't include China, but we could guess (perhaps wrongly) that there's not serious money being made by internet companies there yet. The list also only includes websites, which does not correlate to "internet businesses" (They freely admit this problem in relation to B2B, but what about something like Blizzard?). The restriction to the stock market is necessary for the reason they give, but massively narrows their sample based on an awkward criteria (Facebook isn't listed and is making $100 million or something) 21 companies does not have much statistical significance against seven criteria. I think (I'm not in stats) you'd need more like 600, and then I'm pretty sure you don't start with a list and see if they fit (if you start with a larger potentially overlapping set of business models then you could run a series of algorithmic tests to see which subset provided the best fit). Their list looks somewhat generic, although even at a pinch it looks incomplete. What about selling real goods on subscription? This model has been viable enough offline, and in a way that should translate to online (eg magazine subscription, regular home care product delivery) but it doesn't fit into any of their models.