Web Analytics Opt-out: Is It the Beginning of the End of Web Analytics?

by David Penz on July 15, 2010

David Penz

David Penz

When Google announced on its blog that the company was releasing a browser add-on that gives Web surfers the choice to opt-out of being tracked by Google Analytics, a few confused marketers commented that they thought this would lead to the end of Google Analytics as we know it.

Why would a marketer not want Google to release an opt-out process?  Here are a few of the reasons:

Accuracy of data may decline: If enough people use the opt-out add-on, the number of visits tracked and reported could become significantly different from the actual number. Smaller companies may find it difficult to segment their data and still maintain high enough volumes of visits to maintain consistency in the accuracy of metrics.

Web will lose conversions: People using the opt-out may be typing in your landing page from an offline ad, or for some other reason may have performed some value adding action that you have now missed. When attributing revenue, Web could lose credit since the visitor wasn’t tracked.

Others aren’t doing it: Other Web analytics tools, especially server log-based systems, don’t offer opt-outs. These alternatives are now becoming more attractive to marketers who want as accurate data as possible.

Google Analytics

Data isn’t private anyway: The data that Google Analytics users see is fairly anonymized. Data is reported at the aggregate level, and no personal information is available on any individual Web user. The offering of an opt-out might give some users the false impression that you are actually collecting and using personal information when, in reality, by using Google Analytics you do not have access to personal information. It may also undermine the privacy policy and system you already have set up, which may include your own version of opt-out.

However, there are a few reasons why marketers should not be upset:

It gives consumers choice: Concerns over data privacy grow every time a story breaks that a company like Google, AT&T, or Facebook has had its stores of personal data compromised. These concerns can cause the wary and cautious to avoid websites they do not trust in favor of a website or offline store that they are comfortable with. Giving people the choice not to provide data to you may make enough of a difference in their comfort level to prevent abandonment from a goal or conversion funnel.

It protects against legal threats: Governments around the world are stepping in and passing laws and regulations to protect the right to privacy. Adding an opt-out process to your privacy policy adds one more level of protection against potential legal threats.

Offline tracking isn’t perfect either: Marketers are already forced to accept the fact that they cannot properly track 100% of individuals being marketed to through other channels. For example, if someone orders from a different address than you sent a direct mail letter, to or forgets to provide a promotional code when ordering, their order may not be attributed to the proper marketing campaign. The loss of the ability to track some people that are using the opt-out should also be acceptable, provided the issue is properly explained and accounted for when reporting campaign results.

Data will remain precise: Web visit data will remain consistently measured over time, meaning even though the raw metrics may be lower than reality, you will not lose the ability to compare metrics across segments or evaluate trends over time.

Do you love or hate Google’s decision to create the opt-out add-on? Would you like to see other web analytic tool vendors follow suit and provide similar opt-out tools directly to consumers, or would you prefer to avoid this path?


Google Analytics Blog

About the Author:

David Penz is a Marketing Analyst with SIGMA Marketing Group.

{ 2 comments… read them below or add one }

Supreme Tech Ninja July 24, 2010 at 12:38 am

I don’t see how advertisers wouldn’t be upset. The whole pitch behind analytics is that its easy; insert a little code and get the same data as the server-installed packages. Now it has become “insert a little code, and maybe get some of the same data”. The truly paranoid already have ways to block google analytics from collecting data, and I think the existence of an opt-out (if its existence gets enough attention) will only serve to make the average joe mistakenly think that a greater level of his personal information is/was being tracked without his knowledge that what is the reality. The user ip address can be used to geolocate someone (sometimes) to a general geographic region, but to my knowledge that is as detailed as an advertiser can get when examining an individual user.

On the flip side of the coin, I’m sure Google is trying to stay ahead of the storm that has been brewing around privacy issues, especially with the talk about enacting legislation to deal with such issues. The larger concern for those that think about it will be the fact that having your information being transmitted to a central location (google) means they can track you almost anywhere. If you remember the AOL data leak, it has already been shown how BAD that can be. User X with ip address Y will probably not want anyone, even google, tracking the fact that he went to http://www.myname.com?name=bob_jones&address=123_foo_street and then immediately after went to http://www.icheatonmywife.com.

So I guess what im saying is that although I can see why user’s might want to protect themselves from google having too much of their information, their primary point of concern should probably not be analytics. Being able to opt-out of analytics only makes it more difficult for web site owners to improve, and the fact that now this data that many have come to rely on is (or could be) skewed should justifiably make webmasters upset.

David Penz July 27, 2010 at 4:10 pm

Supreme Tech Ninja – I think you made an excellent point about how this might fuel paranoia and make people think more personal information is gathered than actually is. Google already has taken pains to ensure that Analytics is set up such that data is presented at an anonymous level, and it’s even in the terms of service that you cannot use the tool to extract personally identifiable information, but nonetheless the opt-out may cause people to perceive the opt-out as a ‘fix’ to a privacy problem that may not even exist.

However, I don’t think webmasters and marketers should be as upset about the fact that they will lose the ability to track visits on a few percent of total visitors (unless it turns out that a large percent of web surfers end up adopting and using the opt-out). In the offline world, we are already used to accepting a certain loss of trackable data as long as we can explain the loss. Take direct mail as an example, names and addresses on the mail peice are often matched to names and addresses of customers to identify who responded to a mail campaign, but things like misspellings can lead to an actual response not being identified as such. As long as web metrics remain precise and consistently measured, and we can explain how metrics are affected by the opt-out, we’ll still be able to measure trends and use segmentation to answer our business questions.

It may become important for Google to be transparent about how many people download and use their opt-out and to do some research into how data it provides in benchmarks changes after the opt-out. This would at least give us some comfort that we know how the opt-out is affecting us, uneasiness with the opt-out may be at least in part due simply to uncertainty.

Leave a Comment

Previous post:

Next post: