Google Conversion Room Blog Tips on tracking and improving conversions online

Greater choice and transparency for Google Analytics

Tuesday 25 May 2010 | 18:07


Many website owners use Google Analytics and other web analytics tools to make business critical decisions about how to improve their websites by understanding how users engage with their webpages. 

We’ve worked hard to make Google Analytics both a robust and reliable web analytics platform while also ensuring the trust and privacy of visitors of these websites. Today we’re taking additional steps to provide even more choice and transparency for both website owners and users.

First, we’re announcing the availability of the Google Analytics Opt-out Browser Add-on. The opt-out provides users with a choice of whether information about website visits is collected by Google Analytics. The add-on stops data from being sent from your computer when you visit websites that use Google Analytics Javascript (ga.js) to track usage. The beta version of the opt-out that we are releasing today is available for Internet Explorer, Firefox, and Chrome, and can be downloaded here. 

Additionally, we’re also releasing new functionality for website owners to provide an additional level of privacy for visitors to their sites by offering an option to anonymise IP address information sent to Google. Google Analytics uses the IP addresses of website visitors to provide general geographic reporting. Website owners can now choose to have Google Analytics store and use only a portion of this IP address for geographic reports. Keep in mind, that using this functionality will somewhat reduce the accuracy of geographic data in your Analytics reports. You can learn how to implement this within your existing Google Analytics installation on the Google Code site here.  

We’re committed to Google Analytics as an industry-leading web analytics platform that also provides users with transparency and choice when it comes to privacy.

Posted by Jesse Savage, Google Analytics Team


incrediblehelp said...

I don't understand why you guys are offering this. The whole idea of us using your product is to collect data for our website visits. Why would you allow visitors to stop that from happening?

Anonymous said...

"somewhat reduce the accuracy of geographic data in your Analytics reports"

Why are you making it more difficult for webmasters to track site usage?

Without reliable data you can't make informed decisions.

Anonymous said...

The first paragraph contradicts the general theme of the opt out.

You are basically saying:

We know you use analytics to make business critical decisions, so we've decided to give you less accurate data to base those decisions upon.

webalytics said...

Nice tool - but since it has to be installed first, many people will not use it. What about a more cnetralized approach? Check out our free tool OptMeOut:

max said...

i thing that if you publish a tool and if it has to n#be installed. then it will be pointless because no one will bother to do it so that should be stopped.

Ellie K said...

I so totally agree with the first three comments. I understand Google's motivation in wanting to be fair and demonstrate their interest in protecting the individual's right to privacy. However, by the very nature of web analytics, if site visitors don't want to participate in marketing data collection, then there won't be much use for Web Analytics anymore.

Realistically, most site visitors are going to say, "yeah, sounds good, I don't want anyone collecting my info! No cookies nor Javascript for me!" when given the choice. It is all too easy to forget that there are plenty of downsides to that choice. I became familiar with many of them during my month-long experiment of opting out of everything I could, no cookies in the browser, disabled javascript. It greatly diminished my W3 experience. I would think VERY carefully before ever doing that again!

Returning to the other point: does Google anticipate these sort of opt out features to have an impact on current web analytic algorithms and methods? If so, what recourse will there be, other than modeling using smaller data samples? Or maybe you all are way ahead of me, with a strategy and alternative metrics as WIP? I think the world of Google tech staff....