Saturday, September 22, 2012

Paper: The Use of DPI for Behavioral Advertising (Case Study: Nebuad and Phorm)

From time to time we see the return of Behavioral Advertising as a business opportunity for ISPs, despite the loud crash of these services few years ago. 

Recently I reported on several deployments - in NigeriaTurkey, SaskTel, Marriott Hotels, Orange France, Romania and more. Today's deployments are usually based on an opt-in service, a very different model from the "we dont tell you what we do" model that failed the earlier players - Nebuad and Phorm (which is still there - see "Phorm's New Chinese Operation Valued at $155M" - here and behind some of the above deployments). 

A new research by Andreas Kuehn and Milton Mueller (pictured), Syracuse University, School of Information Studies, examines the political issues behind this service.

This paper examines the use of deep packet inspection (DPI) in online advertising, and analyzes the effects public pressure, regulatory actions and judicial and policy-making proceedings had on those deployments. The research is part of a larger project on the effects of DPI on Internet governance which is funded by the U.S. National Science Foundation [see "NSF Funds Research on DPI" - here]. DPI, which allows Internet service providers (ISPs) to monitor the content of data packets in real-time, can be considered a disruptive technology because of the way its use conflicts with pre-established principles and norms of Internet governance. 

In this comparative study, we examine the rise and fall of NebuAd in the U.S. and of Phorm in Europe. We also include some less visible companies and spill-overs to Brazil and South Korea. We conduct a comprehensive analysis of these cases – from the early development and secret trials of the technology to the regulatory actions, business failures and litigation in the aftermath. Looking at a timeline of several years that covers the dynamic technical, economic and institutional interactions at play, the framework contrasts distinct actors, actor constellations and modes of interaction across institutional settings to illustrate similar and divergent policy outcomes. This research is based upon comprehensive analysis of political and legal documents and a series of interviews with DPI vendors, Internet advocates, engineers, and advertisers.

The narrative follows four stages that we have found in similar case studies of DPI deployments: 1) unilateral, secret deployment, 2) uncontrolled public disclosure of the deployment, 3) civil activism around net neutrality and privacy norms, 4) political, legal and regulatory proceedings to resolve the conflicts. This framework highlights the interaction of technical, economic and institutional factors that are at work when politically contested technologies with a disruptive potential are deployed on the Internet. In this case, as in many others, the analysis shows how the deployments ran afoul of established principles and expectations and how the “notification” and “consent” practices so crucial to privacy law failed to bridge the gap between the expectations of Internet users and the formal legal definition applied by the courts. We show how this gap led to intense political pressure and market exit of DPI-based advertising platforms in both countries.

See and download from "Profiling the Profilers: Deep Packet Inspection and Behavioral Advertising in Europe and the United States" - here

No comments:

Post a Comment