Two presentations at the recent IETF meeting in California underscore the way the Internet’s architecture is being shaped increasingly by advertising-driven content distribution networks.
A research report by Craig Labovitz of Arbor Networks finds that this has major effects on the basic structure of Internet traffic flows: “the majority of inter-domain traffic by volume now flows directly between large content providers, data center/CDNs and consumer networks.” Those conclusions are echoed by Bill St. Arnaud’s recent paper on what he calls “application content infrastructures” (ACIs).
In the same same Thursday session, Balachander Krishnamurthy of AT&T Labs examined how information related to individual users is aggregated by these increasingly concentrated networks as they browse seemingly unrelated Web sites. Thousands of Web sites across numerous categories, countries, and languages are connected through advertising intermediaries which collect and aggregate information about online behavior via the use of cookies.
These two phenomena are related, of course. Internet-based services exhibit a growing reliance on advertising as a revenue source, and as this happens a growing proportion of total expenditure on advertising is drawn to the online world. The growing reliance on ad revenue places a premium on the possession of detailed information about Web users and what they do online. Of course, the majority of users are uncomfortable with, if not downright hostile to, the online surveillance required to aggregate and deliver behaviorally targeted ads. (See the survey by Turow et al, in 2009) These observations point to an intensifying tension between economic pressures and user privacy.
Slides from these presentations were available on the web a while ago but seem not to be now. Labovitz made a similar presentation to NANOG 47 a few months ago which is available online.
Hi Milton, “follow the money” seems (as usual) to be a good way to understand technological developments.
Do you have the complete reference to Turow's (Joe?) survey that you mention in the post?
There is now a link in the article. Thanks for noting the omission.