I didn’t get a chance to write about this when it first hit my inbox, but I just can’t resist saying something about the ridiculous “study” that a consulting firm called Precursor did of the bandwidth that Google supposedly uses but doesn’t pay for. The headline on the email I got — which I assume was sent to tens of thousands of others as well — was sensational and gripping, in the same way that supermarket tabloid headlines are often sensational and gripping (“Elvis clone lands on the moon!”). The email trumpeted the fact that “Google uses 21 times the bandwidth that it pays for.” Bound to get a reaction, right? And it certainly did, with the scholarly-sounding Precursor study being cited holus-bolus by a number of websites.
The fact is, however, that the study has about as much basis in fact as the News of the World headline about Elvis’s clone landing on the moon. As far as I can tell, Precursor took some numbers about market share, extrapolated from that to produce a wildly-inflated bandwidth estimate, and then multiplied that figure by some imaginary number from telecom-lobbyist land to come up with its shocker of a headline, and a press release about how the company “used 16.5 per cent of all U.S. consumer Internet traffic in 2008” (it makes no sense to say that Google “used” traffic, of course, but that’s the level of argument we’re dealing with). Precursor said that Google only pays $355-million for bandwidth, or 0.8 per cent of the total amount Americans pay for Internet access, which is where the 21 comes from.
There are so many errors of logic, fact and simple intelligence in the Precursor release that it boggles the mind. First of all, Google may account for a lot of Internet usage, and what the company pays for bandwidth may be a small fraction of what Americans pay for their flat-rate Internet accounts. So what? The two have very little to do with each other. Is Google supposed to pay more because I use Google image search a lot? As Tech Liberation Front points out, the Internet is paid for by all participants — me with my Internet account, Google with its bandwidth connection, and so on. Asking Google to pay for what I do with my Web connection would be (as Cord Blomquist at TLF says) like asking Best Buy to pay for all the gas I use driving to buy products at its stores. It’s nonsensical.
As Google pointed out in a response on its policy blog, most of the figures in the study make no sense, not to mention that “To say Google somehow ‘uses,’ consumers’ home broadband connections shows a fundamental misunderstanding of how the Internet actually works.” The company also notes that Precursor founder Scott Cleland is paid by the telecom and cable companies to lobby on their behalf against the concept of “net neutrality,” and Mike Masnick at Techdirt notes that the latest study isn’t the only example of Cleland’s problems with math.
I think — as Julian Sanchez at Ars Technica does — that those industries should get another lobbyist, because Cleland’s study is about as persuasive as a briefing paper prepared by Big Bird and the Cookie Monster.