Of course, that type of research generates a lot of "touch costs" that make a large number of survey participants very costly. By automating the process through the Internet, many of those costs can be eliminated. That way, companies like ComScore can have a large group of Internet users that they can sample in different ways, depending on the survey being performed. The result should be better market research.
Unfortunately, it's also very easy for this automated process to lack clear disclosure to -- and the explicit consent of -- the people being monitored. Within the past month, Computer Associates' Benjamin Googins and Harvard professor Ben Edelman have both pointed out disclosure and consent problems in ComScore's software.
The most disturbing thing about this situation is that it's not new. ComScore has been in the middle of this controversy for at least five years, and very little seems to change except the name of the software. The company claims that it does notify its participants, yet researchers like Googins and Edelman clearly document how a user could install this software without the foggiest idea of what it's doing. IT managers at several organizations have blocked ComScore software because of its intrusive habits that include rerouting secure HTTP traffic containing credit card information.
As Larry Dignan notes, there are serious questions about the quality of ComScore's traffic measurement strategy, perhaps because of the shortcuts the company has taken to recruit participants. Everyone that runs or advertises on an Internet site wants better information, but it doesn't seem ethical to gather that information off a user's system without clear disclosure.