Archive for March, 2009

AdSense publishers: we’re all in this together

March 25, 2009

 

Google published some clarifications for AdSense publishers on the new interest-based targeting program.  This one caught by eye:

However, even if you choose to opt out of displaying ads based on interest categories now, you’ll still need to update your privacy policies by April 8th. This is because your site may still show ads based on a user’s previous interactions with an advertiser, such as past visits to that advertiser’s site.

 

via Inside AdSense: Answers to your interest-based advertising questions.

It makes perfect sense, and is interesting to note that, as a publisher, you’re in the program at some level no matter what.  Will Google really enforce this come the firs week of April, which is to say, limiting the kinds of advertising shown on sites that have not adopted the new privacy policy or withholding checks?

Google Watches Its Language via Mediapost

March 20, 2009

Google’s rhetoric carefully avoids use of “tracking,” “following,” or even “tagging” or “collecting” to describe this process. Google “remembers” your visit. Your cookie, by the way, is also humanized, because in Ghosemajumder’s construction, it talks to Google but never “tells” Google who you are. Most of the time he does not even talk much about what is being stored on your cookie on your browser. Instead, he frames the process as putting your cookie into an interest category. Interestingly, what started as a cookie that Google “stores on your browser,” quickly becomes “your cookie number” and then becomes “your cookie.”

Just as interesting are the arguments that are not here. At no point is he suggesting that the ads pay for your free search services or even that there is a fair exchange of value. Google is not bartering with us. It is presenting itself as a partner in an effort to give you more relevant information. What is interesting to me about the pitch is how he elides publisher, consumer and advertiser interests pretty fluidly rather than suggesting any kind of natural conflict of interests, any intrinsic relationship that might require a barter, trade or negotiation.

via MediaPost Publications Google Watches Its Language 03/20/2009.

The roll out interest-based targeting was a PR masterstroke.  All you have to do is look at the headline from the NY Times coverage.

capture 

The B-word doesn’t appear in the headline — I see high-fives in the Google PR control center.  They did their job well.

What’s behavioral targeting worth?

March 20, 2009

 

It’s very hard to tell at this point what behavioral targeting is worth, but here’s one indication of the willingness to pay for it:

…this year, advertisers on average would pay 10 percent more for both demographic targeting and daypart targeting; they would pay 13 percent more for behavioral targeting.

 

 

via Search report shows obsession with behavioral targeting | Blog | Econsultancy.

No surprise:  there’s a lot of money at stake when it comes to tracking individual preferences, and strong forces for the adoption of profiling.  If consumers own their profile, how do they get a piece of  that value?

Google’s Behavioral Ad Beta Scares Publishers Over Contextual Relevancy

March 19, 2009

 

It is important to remember that it is still up to publishers whether to participate in Google’s behavioral targeting network.  (It’s opt-in for publishers and advertisers, opt-out for consumers.  Hmmm.)

The concern is that Google’s contextual relevancy may dwindle down, in exchange for behavioral relevancy. Contextual relevancy is what Google takes pride in, they love to say they have ads that match the exact content of the page. If you are reading about Sony video cameras, Google will show you AdSense ads for Sony video cameras. But possibly now, if you are a sports enthusiast (Google knows this because you visit sports sites) and you are reading an article about video cameras (because you want to buy one for the next game), you may see sports ads and not video camera ads.

 

 

via Google’s Behavioral Ad Beta Scares Publishers Over Contextual Relevancy.

PS This reminds me that I want to track the number of documents indexed in Google that match Google’s privacy policy as written here.  That’s one way to measure adoption of the new program by publishers.

First day results (and the road ahead)

March 17, 2009

Yesterday, Monday March 16, was the first real day of any promotion of the privacychoice.org opt-out wizard, and here are the results:

Last time I checked, over 325,000 opt outs served.  We don’t keep detailed user analytics, but by my estimate, that’s something like 7,000 interactions for the opt-out button.  Over 50 people have subscribed for email updates, and more than 60 people are following privacychoice on Twitter.  

All this without any mainstream news coverage — just from commenting on blogs, users recommending to users, and a few links on some high-traffic blogs.  It seems clear enough that consumers want a choice when it comes to online profiling, and that the opt-out wizard is a significant — though flawed — first step.

A great debut, but there’s much more work to be done.  Here are my priorities:

1.  We’re moving everything over to privacychoice.org, which is the long-term URL and name for the project.

2.  Later today we will push a minor upgrade to the site, mostly improving the words, and eliminating ShareThis (which made some users uncomfortable due to third-party cookies).

3.  Browser plug-ins are in process, after a deep think on the approach (will post on that later today).  I don’t expect to see nearly the same uptake as the one-click button, but it should be appealing to more experienced users.

4.  We have a plan for some quick improvements in how the wizard operates, to make it much faster and also to skip sites that don’t respond in a timely way (so browsers don’t hang there).

Thanks to everyone for a great first day.  Don’t forget to tell a friend, and to send us your comments and suggestions!

The third-party cookie compromise

March 16, 2009

Here’s an important comment we received over the weekend under the headline, “We don’t set cookies?”:

No, but you do allow others to set cookies via your front page, namely .sharethis.com, which you will spot has a preceding “.” meaning that all subdomains have access. I think its disgraceful that a site pretending to be pro-consumer choice is far from it. My view is that “privacychoice.net” is a site that from the outset intends to mislead its visitors. If your really anti-tracking, then dont allow ANY cookies on your own site.

This highlights an important issue — can privacychoice.org use third party cookies to increase the effectiveness of the site?  In what circumstances and with what assurances?

As our FAQ points out (see last question) we include third party cookies for ShareThis (which allows users to help promote us in social networks) and UserVoice (which allows us to collect and respond to suggestions).  I reviewed the privacy policies of each provider, and determined that neither of them is currently using any advertising (so no additional third parties would be setting cookies) and that they use cookies now only for the purpose of delivering their service.  

Nevertheless, my plan is to move away from using ShareThis as soon as possible, since ultimately their business model is likely to include harvesting cross-site information one way or another.  UserVoice, on the other hand, should thrive as a subscription service (and it helps that they have an office right down the hall, which increases their accountability).

This is a great example of the kinds of choices that face any online publisher — when is it okay to expose my users to third party cookies in order to make my service better, or to support it financially?  I don’t think it’s enough just to say, never.  For me, I drew the line ultimately at Google Analytics — it would be enormously helpful to have Google Analytics data in order to improve the experience for the users of privacychoice; but I chose not to because I’m not convinced that that the information Google receives in that process will always be used in a manner consistent with user expecations, particularly for a site like privacychoice.

I suspect this is the first of many difficult compromises for this project.  My goal is to be completely transparent about these questions, and to use this site as an example of the kinds of real tradeoffs publishers must make when they live, as we all do, in Google’s world.

We’re on to something

March 16, 2009

Well, today was really the first day of any promotion for privacychoice.net, and the traffic has been terrific — hundreds of users used the one-click process today to opt out of behavioral targeting across all of the big ad networks.

counter

I’m delighted by the organic growth — since we’ve only had two real blog placements in the last 24 hours, it seems like the service has viral appeal.  

This week:  migrate over to privacychoice.org … and hopefully launch our plug in … stay tuned!

Our first blog mention

March 15, 2009

 

For me, this is sort of like the dollar bill in a frame on the wall:

Internet Advertising: Opt-Out Of Behavioral Targeting.

And we’re seeing nice growth in the opt out number, every time I check!

Some G2 on the Google Privacy Plugin

March 15, 2009

 

Google is offering a plugin that keeps the optout cookie on your computer even if you clear cookies.  This thoughtful chap took a look at the source code and reports:

After examining the source however, it seems that all the plugin does (at least in Firefox) is check to see if the cookie is there when you load the browser. If the cookie is not found and third party cookies are not disabled, then it adds the opt-out cookie.

via The Sitting Tree » Blog Archive » Google Behavioral Targeting Ad Program

It will be cool to connect the opt-out db with this.

(Hoping (in sort of a lazy way) that the component that checks  for the presence of the Google cookie may be a sufficient indicator of whether to run the entire sequence of opt outs.)

UPDATE:  I think we’ve learned now in our own handling of the Google cookie that it doesn’t look at all at cookies that are present.  It simply re-writes the hardcoded opt-out cookies upon certain events (new session, cookies being cleared, regular interval).

Soft launching privacychoice today!

March 13, 2009

After a bit of testing and a few rounds on the interface, I’m pushing privacychoice.net out today to get some feedback.  Given the buzz about Google’s new behavioral (er, interest-based) user profiling, the timing seems good to get some attention to this effort.

Check it out and let me know what you think:

http://www.privacychoice.net

PS In a day or two we will switch everything over to privacychoice.org, which is the long term home for the project.