Every once in a while, I come across some really bad conclusions based on some really bad research. The A First-Ever Research Study: Estimating Google's U.S. Consumer Internet Usage & Cost -- 2007-2010, from netCompetition.org, is one such document. Promising to be "straightforward, open, transparent, and replicable as possible" in its methods, the group's research findings are based on a mishmash of sources randomly aggregated to a wholly unfounded conclusion that Google isn't paying its fair share of bandwidth.I don't know if Google or any content provider is using more than its fair share of the Internet and, frankly, it would be interesting to find out how the Internet is being used, but taking usage data from a Cisco report, a report from the USPTO peer-to-peer file sharing, search engine and video-sharing market share data and a few others, and then coming up with some percentage of total Internet usage, is wildly unreliable. What exactly is the relationship between market share and bits per second consumed?

The response of Richard Whitt, Google's Washington telecom and media counsel, to this so-called research is that that the author, Scott Cleland, is wrongly equating consumer's use of Google's services such as Gmail, YouTube, and the search engine as Google originating traffic. Whitt's right. You, the consumer, who already has paid for your bandwidth and access to the Internet, is using content hosted at Google, or Yahoo -- content providers who also have paid for their bandwidth and connection to the Internet. Neither Google nor any content providers are forcing content on you.

Consumers pay for their access. Content providers pay for their access. The telecoms, in providing the access to consumers and content providers, already are getting paid to carry the bits by both content providers and consumers. That's fair.

Through peering arrangements, (you can read more about peering, transit, and transport in this paper), some tier 1 service providers such as Verizon and AT&T agree to exchange traffic essentially for free because that lets their customers, content providers and consumers alike, interact. The traffic that flows between tier 1 providers is roughly equal and the peered ISP's have an equitable arrangement. Peering is similar in nature to reciprocal charges in the public switched telephone network (PSTN) where each telcom company paid a few pennies to have its customer calls terminate on another telecoms customer. Lesser service providers pay transit or transport costs to their upstream ISP's because the smaller providers are getting more benefit -- access to the Internet -- than they are giving. Its how the world works. Or doesn't work, like when Sprint cut off Cogent last October.

However, the tier 1 ISPs that form the backbone of the Internet, also carry traffic where neither the content provider nor the consumer are paying that ISP. Those bits are transported for free. But that works out because the tier 1 ISPs also get their bits transported for free by other tier 1 ISPs. It's still a fair and equitable arrangement and has worked out pretty well.

But what the ISPs really want is a chunk of the advertising revenue that content providers receive, and that advertising revenue is a substantial sum. Look at the size of Google, Yahoo, and other ad-based businesses. The various attempts by some ISPs to replace the content providers' advertising with ads that pay the ISPs instead shows the lengths that some telecoms will go in tapping into that revenue stream. Alternatively, ISPs also could charge content providers like Google and Yahoo again to move the same bits over the same pipes that you and the content providers have already paid for!

This is going to get ugly. Real ugly. Big money is at stake and you can bet someone like the consumer is going to get trampled "for our own good." Too bad the ISPs can't figure out how to make money by pushing bits.