Why Not Track? Results from an informal survey

Recently Krux Digital (a PrivacyChoice partner) and Mozilla published interesting statistics on the adoption of Do Not Track preference in Firefox. They suggest that so far over six percent of Firefox users who have the Do Not Track option have now enabled it. As Mozilla has pointed out, these elections exceed the number of people using privacy add-ons (like AdBlock Plus), and it dwarfs the number of people who have found and used privacy options through the ad industry initiative to place notice icons in ads.

Given significant early adoption, it’s important to understand why people are making this choice and what they think it will mean to them. To get a sense of this, we’ve been running an informal survey on privacychoice.org, which is offered to users who visit with the Do Not Track header. These users see an intercept like this, which allows them to click to the survey.

Results. Here are the results based on just over 100 complete responses:

Implications. This survey is unscientific (see below), but the results warrant more research about user attitudes about Do Not Track.

These points were notable:

• Most users went looking for it. More users heard about the new choice and sought it out than those that stumbled across it while in their privacy settings. What does that suggest about find-ability of the feature?
• Most users care about data collection. Controlling data collection is far and away the most important motivation for the election.
• Some users had misconceptions. A significant number of users believe that this choice will limit the number of ads they see.
• Most users are realistic. These early adopters are more aware than I expected about the limited practical effect of their choice, since only a handful of tracking companies honor Do-Not-Track signals.

Shortcomings. This research is informal, preliminary and non-scientific. Problems with this study include (but are not limited to):

• Small sample (only around 100 respondents);
• Self-selecting sample (we expect PrivacyChoice users to be more “privacy aware” and “privacy concerned” than typical users, and they may have previously learned about Do Not Track through our service);
• Unscientific survey construction (we randomized the order of responses and such, but I’m no survey scientist); and
• Ambiguity in questions and answers.