Google announces plans to offer Google Analytics Opt-Out

I was on client calls most of the day, and when I came up for air in the afternoon to check my RSS reader and Tweetdeck to see what what going on in the world I made a fascinating discovery.  Like many of you I came across the following post from the Google Analytics Blog:

http://analytics.blogspot.com/2010/03/more-choice-for-users-browser-based-opt.html

This was most unexpected, and my Thursday suddenly got alot more interesting.

Before we go any further let me state that I have not been briefed by Google on this news item.  This post is purely based on my own initial thoughts on the matter.

The blog post announces Google's plans to release a browser plug-in that would allow consumers to opt-out of Google Analytics tracking.  This offering is still in development, and the post offers no specifics on the release date, although it implies that this is only weeks away.

(Side note: It is also interesting to note the language used in the post.  The post leads with "As an enterprise-class web analytics solution..."  This isn't a surprising or entirely inappropriate assertion, but it strongly implies Google's aspirations for GA.)

There are many reasons why Google's course of action is counterintuitive.  Naturally, the marketer in me recoils at the idea of voluntarily allowing measurable data to slip through our hands.  Rationalizing web analytics data is already hard enough, and now this?  And we can certainly debate the true privacy impact of web analytics on consumers. 

However, it is safe to assume that the bright people at Google aren't doing this on a whim or because they are bored and looking for something to do.  They take Google Analytics very seriously and are quite deliberate in its development.  But Google treats their web analytics offering very differently than a traditional enterprise software vendor might.  Google Analytics sits within the unique context of Google's other businesses, namely it's search and advertising offerings.

Don't panic.  It is highly doubtful that this will do substantial harm to Google Analytics and its customers. Why?

  • Actual opt outs will not deal a death blow to data gathering.  Although some people will undoubtedly use the Opt Out plug-in, most will not. And even if the numbers drop a bit, the trending data will remain valuable and web analytics experts will learn to quantify and manage any resulting bias.  This may be analagous to cookie deletion and enabling javascript on browsers; web analytics has survived despite the fact these things can be easily deactivated because relatively few consumers actually take advantage of those options.
  • Good privacy management drives Google's business and profits in the long term.  Google Analytics adoption may actually increase in the wake of the Opt Out plug-in, particularly outside the United States.  This move shows good faith to regulators and sets the stage for Google Analytics to operate safely within the containts of more stringent privacy regulations in Europe, which may boost GA usage and ultimately drive more advertising revenue.  Again, GA operates within a larger context at Google, and they have to consider privacy and other regulatory concerns as a material risk to the business if they are not managed properly.
  • Opt-In consumers are (slightly) better customers.  Many other offline and digital channels have adopted standards for preference management and survived. The marketing axiom has always been that opt out provisions make marketing more efficient and more relevant becuase you no longer waste resources on consumers who aren't interested in your company or products.  This comparison won't hold up 100% in the case of the web, because the marginal cost of serving additional visitors to a website approaches zero.  However, your site's targeting was probably falling flat for the consumer who is inclined to opt out, so focusing your efforts on those who are willing to be measured may enhance the achievement of site goals.

It's not the end of the world, but there will be an impact.  Here are some things to start thinking about ahead of the Opt Out plug-in going live:

1. Be ready to benchmark your traffic and known visitors before and after the release of the Opt Out plug-in, so you can evaluate what, if any, impact this has on your traffic measurement, and whether the traffic dropoff is correlated to the visitors who actually purchase or meet your site's goals.

2. Be ready to accomodate the bias that GA opt outs incur on your metrics and recalibrate segmentation and testing strategies if needed

3. Depending on your business, there may be an opportunity to use the analytics Opt Out phenomenon as a way to build trust with visitors. Your use of Google Analytics could be positioned as a clear indicator that you respect visitors' privacy, and sets the stage to invite them to submit more information about themselves (alongside a clear and readily available privacy policy) to get increased value in a transparent way that visitors fully control. 

4. Expect that reconciling website metrics with other data sources for attribution will be more difficult now that more of the visitor data is missing.

The other wrinkle to this news is to consider how this will impact non-GA users.  Will site operators who use web analytics solutions other than Google Analytics feel pressure to match the privacy capabilities that GA users offer to their visitors?  Can this become a differentiator for online businesses?

And for the rest of the web analytics vendor community, if Google's move pays off, will opt out become a standard that other vendors feel compelled to follow?  Do we need a global standard for a web analytics Opt Out protocol?  Would a cottage industry of brower based opt out plug-ins develop overnight?

What do you think?  Let me know!

Comments

Google Analytics opt-out offering

Joe,

You got close to the reason Google is offering the opt-out in your second point above about usage outside of the United States. I think, and I have not been briefed on the subject either, that Google is setting themselves up to more or less take over measurement within the U.S. Federal Government. Did you see this blog post from a few weeks back:

http://analytics.blogspot.com/2010/02/powerful-flexible-secure-and-now.html

If, for example, Peter Orszag and the Office of Management and Budget (OMB) were to be considering re-writing the rules for cookie-based data collection, and if, for example, there were a different set of standards for "free" versus "fee-based" solutions, and if, for example, parts of the Fed were already pretty cozy with Google's CEO then this move would make incredible sense. Especially if one of the requirements in the OMB's new rules was a permanent opt-out of tracking to placate the privacy folks (EFF, etc.)

What do you think?

If my thesis is correct this is the beginning of a whole new area of expansion for Google Analytics. Considering that most Federal agencies toil away under outdated software-based implementations, and that Google Analytics comes with the "deficit busting" price tag of ABSOLUTELY FREE, my suspicion is that GA is poised to become the "new standard" for web analytics within the Federal Government.

Otherwise I personally think this is a non-announcement, more a check-box or hoop-jumping exercise for Google than a threat to our data, other vendors, or web analytics in general.

I'd love to hear your thoughts!

Eric T. Peterson
CEO, Founder, Senior Partner
Web Analytics Demystified, Inc.
http://www.webanalyticsdemystified.com

google analytics

Right now most other leading companies such as Omniture or Web trends already offer an analytics opt-out, albeit cookie based. Google recently began offering a plug in that allows an opt-out of behavioral ads, to enhance the cookie web based opt-out DoubleClick long provided, so it was behind on this issue. Now that it did decide to offer an opt-out, it could hardly offer only the more limited cookie based version.

As Eric notes, the government factor is relevant (we at the Future of Privacy Forum have also urged OMB to push for an enhanced opt-out)...and a plug in is better than nothing and is a good addition to a web based opt-out. But requiring users to download something in order to make their opt-out choice stick isnt a very friendly solution....either browsers will need to support a "firm" opt-out cookie or other options need to be developed by industry.

If the default privacy impacts of data aggregation about users are to be balanced by providing opt-outs and choices, and that is the policy argument supporting the default - the choices users make need to be respected.

Opt-out rates for behavioral advertising run at about 1%, let's at least make it work right for those who take the trouble to express their concerns.