Back in the 1990s, technologists proposed that privacy be dealt with through “infomediaries,” companies that would negotiate on behalf of the consumer to protect privacy and other rights. For instance, in a 1997 editorial, Esther Dyson argued that infomediaries were superior to self-regulation, privacy laws (ineffective because of the global nature of the internet), and other forms of consumer self-help (because consumers “aren’t very good at protecting their own interests”). She argued:

“The Internet, however, promises to change this balance of power. It can give each individual the power and the tools to bargain on his own terms, even as computers provide the capability to keep track of each individual’s preferences about how the data are to be used. But it won’t if Washington imposes one-size-fits-all regulation.


“The Open Profiling Standard is a uniform way for individuals and businesses to represent personal data and to express their privacy preferences and practices. OPS allows individuals to control which data they reveal to any particular Web site (unlike the notorious “cookies” — data-collection software routines that hide on a consumer’s computer and are unintelligible to anyone other than a programmer). Any user can look at his own OPS profile and make sure it’s accurate and as complete — or incomplete — as he wishes. [1]

Wow, was Dyson wrong. In retrospect, her proposal looks like snake oil—and perhaps for the reason she acknowledged in the editorial: that consumers are too busy to protect their own privacy. But on a larger level, the internet only heightened the problem of adhesive contracts despite the theoretical ability to tailor terms to individual transactions.

Infomediaries would have been the ultimate advertising targeting tool, because presumably consumers would trust infomediaries with their most secret desires. Advertisers could pay into the cooperative and target individuals with great insight. But for it to work, the infomediary would have to have a monopoly on the most valuable personal data. And therein lies the problem. These business models all failed because buyers of personal data could always get “good enough” targeting information from other third parties. This good enough information was less expensive, both in terms of monetary cost, and in fewer privacy restrictions. For infomediaries to work, there would have to be some strong economic or legal force to require marketers to use them and not to obtain information from substitutes such as data brokers.

[1] Esther Dyson, Protect Internet privacy—privately, Wall Street Journal, Jun. 17, 1997.