My RapLeaf Profile is Amusingly Mistaken. This is What the Fuss is All About?
By Eric Goldman
The latest in the Wall Street Journal’s “scare journalism” series of privacy articles is a hatchet job on RapLeaf. I don’t know much about RapLeaf, but at conferences, privacy advocates frequently invoke RapLeaf as an omnipotent, omniscient privacy violator–a stereotype the WSJ article seeks to inflame.
This morning I decided to see what the fuss is about. Rooting around for the opt-out mechanism mentioned in the WSJ article, I discovered that I could see what information they know about me, and I could even customize it. With a little trepidation given its reputation, I logged in to RapLeaf via my primary email account [email@example.com] and this is what I learned (I edited the format but not the substance):
Below is information we provide to companies to personalize the content and advertisements they show you. This information comes from public information you have posted online as well as information from some of our partners (from sources like surveys, census data, and public records).
Companies use this information to personalize content for you. We’ll continue to update this view with more information as we obtain rights from our partners, and as we continue to add new information. You can remove segments by clicking on the “Remove” link to the right and soon you’ll also be able to add new segments. Some of this information may go into a cookie.
Entertainment > Music
Internet > Online Streamers > Photo Sharing Consumers
News & Current Events
News & Current Events > Online News
Shopping > Auctions
Shopping > Online Shopping
Social Networks & Online Communities
Social Networks & Online Communities > Blogging Resources & Services
Social Networks & Online Communities > Business Networking
Social Networks & Online Communities > Journals & Personal Sites
Social Networks & Online Communities > Social Networks
Location: San Francisco, California, United States
Influencer Score: 81-90
Children in the Household: No Children Present
Household Income: 0-15k
Marital Status: Single
Home Owner Status: Own
Your Online Profile
The following is public information we gathered online which we have associated with you. We do not use any identifying information (like a social network id, email address, name, or any other ID) in Rapleaf cookies. The data below is only accessible by companies approved by Rapleaf who are personalizing services to you. You can remove attributes by clicking on the “Remove” link to the right. You can also opt out altogether here.
Name: Eric Goldman
Location: San Francisco, California, United States
Assistant Professor at Marquette University Law School
Law professor at Santa Clara University School of Law
1) The site appears to have partially confused me with another Eric Goldman, as evidenced by the age (wrong by almost a quarter-century), marital status (married nearly 13 years), parenthood status (two super kids), and household income (off by a lot, fortunately).
2) The interest categories are amusingly generic. They aren’t provably wrong (though my wife would guffaw at listing me as a “shopper”), but they don’t really do anything to uniquely describe my idiosyncratic interests. So an Internet user is interested in news, entertainment, shopping, social networking…? Pretty insightful, guys.
3) The listed websites are also laughable. For example, the most defamatory thing in the profile is that it accuses me of having a MySpace account. Uh, no. In fact, the linked MySpace profile is http://www.myspace.com/jodydog12. Huh? I do have a Friendster account I created in 2003 and never used. I don’t know if I created a Multiply account, but it’s content-free. I have several Flickr accounts but it links to the one I don’t use. It does accurately reflect my Twitter, Friendfeed and LinkedIn pages, all of which have lots of accurate information about me–which makes RapLeaf’s other factual errors less excusable.
This RapLeaf profile raises two questions:
1) Why doesn’t RapLeaf do a better job knowing me? The WSJ story has examples (intended to be horrifying) of deep insights into users. Despite my spending 14 hours/day on the Internet, RapLeaf sure hasn’t gotten many good insights into me. One hypothesis: I don’t hang out in the dicier areas of the Internet. For example, a number of the examples in the WSJ article appear to be tied to using Facebook apps, and I categorically don’t use apps. I think Facebook has a serious brand/consumer trust issue with bad behavior by its apps providers.
More generally, RapLeaf’s shoddy profile reflects how hard it is for any web service to know everything it wants to know about consumers. I discuss this challenge in my Coasean Analysis of Marketing article.
2) Should I be bothered by RapLeaf maintaining a profile about me, accurate or not? I am more sanguine about data profiling than most people. As I wrote earlier, “relevancy trumps creepiness.” Nevertheless, it’s embarrassing for RapLeaf to do such a poor job figuring out who I am given how much information I’ve made public. If RapLeaf can’t solve its data quality problem pronto, it seems like the marketplace will take care of it faster than any regulator could.
Even so, RapLeaf remains a scary “privacy threat” that creates moral panics among regulators. Whatever RapLeaf’s fate in the market, I would hate for us to overreact to its existence in developing privacy regulations. I can think of several websites (Google, Facebook and LinkedIn come to mind) that know way more about me–including my email address and real name–that could be adversely affected by miscalibrated regulatory intervention.