There seems to be a bunch of problems with them making a definitive statement about internet business based on this research.<p>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.<p>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?).<p>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)<p>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).<p>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.