Georgina Armstrong

What Kind Of People Are Using Your Website?

by Georgina Armstrong

2013/12/02

Co-author: Catherine van Zyl

QUESTION: What if we said your website was most popular with 25 – 34 year old women who love movies and cooking?

We are super excited about this development in Google Analytics (GA) technology: for the very first time you can get a real idea of what kinds of people are actually using your website. Google officially announced the addition of this new feature on the 1st of October 2013 at the annual partner conference in Mountain View, California. In short, with one quick change to your GA tracking code, you will start to see data such as age group, gender, interests and affinities.

What About My Privacy?

The thought of a company being able to gather such intimate-seeming data can be a bit alarming, so let’s take a step back and talk about the privacy issues. Those of us who use a combination of Gmail, Google+ and search can begin to understand how much Google knows about us, but Google has to maintain a reputation for respecting user privacy. So while they do know an awful lot about you from your Gmail and Google+ accounts, that’s not the information that’s being used here.

What is being used is aggregated data gathered using the Google Display Network (GDN). This is data that Google has collected by setting cookies on your browser when you are looking at sites with GDN banner ads on them. (Cookies are small text files which remember information for you or a website, such as your log in details or whether you browsed that site). The banners set cookies in your browser that say “this user visited websites with articles about cars”. So, when you arrive on a site, that information (“car lover”) is communicated to that brand’s GA account.

The cookies in your browser then send Google interest and affinity data based on your browsing history. Then Google uses an algorithm to estimate how old you are and whether you are male or female. It is essentially a big data-informed assumption that what you look at on the internet gives Google clues as to the type of person you are.

So no, Google isn’t releasing your verified personal information, just who they “think” you might be.

What will be visible in Google Analytics?

This data appears in Google Analytics under Demographic Reports. It’ll show male and female user percentages and how those groups behave on site. This is an example of the data:

 

In this hypothetical example you can see that men and women behave fairly similarly on site. This may not always be the case though. If the metrics weren’t similar brand managers could investigate what might be causing men to bounce more than women. (A bounce is a visit where the user only sees one page and leaves immediately).

 

It is important, for the user’s peace of mind, to take away that you never appear as an individual in Google Analytics reports. All the data is aggregated to protect the user’s privacy. For example, GA doesn’t allow us to see how many 28 year olds visit the same site. Age will bucket in broad age group categories like this:

 

Brand Managers will be able to investigate the behaviour of women aged 18 – 34 using the Advanced Segments feature. Advanced Segments will allow them to draw better conclusions about how different groups of users behave on the site. They could ask questions like “Are my older users struggling to complete a certain action?” or “Are my male users less interested in my content?”

As discussed, by tracking what content you look at, Google’s algorithm forms a generalised profile of who you might be. Here’s how it estimates your interests:

Say you read a lot of content on cars and news over a period of time. Google will assign you the interest categories of “Automotive” and “News/Weather”. 

Later, you start to focus more on world news and travel. These will slowly become interests too as more cookies are set on your browser as described above.

Eventually, Google can assign what they call an “affinity” to your profile. One profile can have many interests and many affinities. An affinity is a description for a user based on their interests profile.

Google may decide therefore from the above scenario that you are an “Auto Enthusiast”, a “Travel Bug” or a “News Junkie” because you now have the right interests to place you in these categories.

This affects website owners because it allows them to tailor their content to their audience. It doesn’t help to write content about cooking when most of your audience is interested in “Arts & Entertainment”. Here is an example of how these new GA reports can help you:

From the graph above, the users are broadly interested in news and audio-visual entertainment. The unfortunate part about this way Google reports this data is that you can’t see any overlap; you can’t tell if person X is a Music Lover, a TV Lover and a Technophile.

The bottom line

We now have a way to actually look at what our user base is interested in. Not just our gut feel but data driven estimates that tell us, with some accuracy, who these people are and how exactly we can be more relevant to them. What’s the ultimate benefit for a brand manager? They can build a content strategy that aligns with your interests and, hopefully, grow their audience. They’ll also be able to track how much the audience they are focusing on is growing and whether they are becoming more engaged with the website!

About The Author

Georgina Armstrong joined the Cape Town Quirkstation in November 2011 as a Senior Web Analyst. Educated at UCT, she went on to do her MBA UCT Graduate School of Business, including a semester at NYU, bringing some global experience to Team Optimise.

Comments

Saw that tab on GA just this afternoon and wondered how that data would be collected. Now I read this. Cool beans!

Posted by Austin on 2013/12/04

Informative article, great.

Thanks.

Posted by Marius on 2013/12/05

Make a comment

To prevent GottaQuirk from becoming spam central, we block the use of certain words like porn, sex etc. We apologise for any inconvenience, but can't spend our lives deleting messages left by spammy friends.

Captcha