This post is also available in: Russian

For efficient decision making based on Web analytics, it is often useful to import additional information on pages, visits and users. Based on this, you can analyze specific sections of your site, content categories, special visits and users of a certain type. All this data reflects specifics of your business and your Web project. Without enhancing your analytical data with your business details, analysis would be quite general and not so effective. Making an economic analogy, basic Web analytics is just a technical tool and rich Web analytics is like insider information, although non-classified. From newsreels you probably know how effective it is. In this article, we will discuss mechanisms of sending advanced data to Google Analytics and the best applicable practices.

Rich statistics in Google Analytics allows you to enhance the data on page views, visits and users with your business specific data.

Custom Variables: Description

To send more details to Google Analytics, you can use custom variables. Each time a page is viewed or event is triggered, Google Analytics data is updated. Each variable has its name, value, level, and must occupy one of the five available slots (hence there is a restriction on 5 variables per page). The values are sent not within a separate call, but with the data on page views (or events), so you need to set the variables in advance. Using the variables of different levels, you can attach rich data to specific pages (page visit), visits (series of page views) or visitors (persistent data saved between sessions).

To enrich analytical data with your specifics, you should design and implement the code to set user variables. When segmenting the data based on page views, visits and visitors, you should determine what attributes might interest you. Define when you need to register the visit/visitor data and send it to Google Analytics. When making your design, consult Google’s native description.

Consider user-defined variables as a prerequisite to Web analysis. If there is nothing special in your site pages, visits and visitors, then most probably your Web site has nothing to do with your business altogether.

Custom Variables: Best Practices

Page Level Custom Variable
Page level variables determine the data specific to a particular page.

When using SEO techniques, it is usually almost impossible to derive page category from its URL. This is especially true when the site contains a large amount of content on different topics. A highly popular approach is to use page level custom variables to categorize content. This mechanism allows you to enhance business intelligence using CMS data. Then you will be able to identify the most popular content category, variations of user behavior in different categories of content, etc.

In addition to lots of valuable information, CMS often stores details on content authorship. This may not necessarily be the author of a blog post; it may be a film studio or data presentation option. For example, if content on the page can be presented in different layouts and markups, it is useful to analyze popularity based on layout. Understanding that a wide picture content layout increases time on the site and the number of conversions (subscriptions) may be seminal for further development of the site and its success (and for your personal value for the company also).

Summing up, we highly recommend you to have a close look at your CMS and identify the data to set segmentation in Google Analytics.

Visit Level Custom Variable
The visit level custom variables reflect specifics of every visit to your site (a series of page views from the landing page to the exit page).

The most common approach to the visit level custom variables (session level) is to distinguish between the logged on and anonymous visits (sessions). This approach allows us to identify behavior of users logged on to the site.

More generally, while planning the session-level variables, we should consider the differences between visits from the business perspective. A classic example is visits achieving a specific result (conversion). For commercial results, Google Analytics offers a special eCommerce tool. In other cases, you’d better consider custom variables. Ultimately, the task of a Web analyst is to say what should be done to improve the business results and determine investment priorities. With custom variables, you can speak the same language with other project participants and develop your recommendations based on deeper understanding of the content of the site and its visitors.

Visitor Level Custom Variable
The visitor level custom variables define user-specific details.

When analyzing the information and making decisions, it is sensible to make separate analysis for certain groups of users. This separation is often dictated by the needs of your business and Web project. For major content portals, users are often classified into common visitors, registered users and buyers of a certain level (i.e. one-time buyers, regular buyers, etc.). Anyway, visitors are your customers, and Google Analytics knows only technical details about them. So you can use your business-specific customer data to enhance your Web analysis. The more you know about your users, the more likely you’ll offer them exactly what they want and are willing to pay for (at least with the time spent at your site).

As seen from the best practices, the main objective of the custom variables is to make Web analysis more clear to the business, so they can use a common language to describe data, visits and users.

Use in Blogs and Content Portals

As a real-life example, let’s discuss the custom variables we have used for our blog.

• Segmenting content by categories. It is convenient to cover various topics on your site. For instance, our blog is dedicated to online media technology, Web analytics and advertising. Each topic is worth a separate site, but we have combined them, as they make up the foundation for most modern Internet projects.
• Distinguishing between the staff and outside visitors. This might be useful at the initial stage, when the resource popularity is not so high and a large number of employee visits can affect Web analysis.
• Segmentation of users into common visitors and subscribers. It is very important to understand behavior of the most loyal visitors.

Unconventional Approaches

Custom Variables can also be used for unconventional, far from intuitive, applications. For example, you can target special offers to users of a specific group, offering a discount for registered users or free video for regular buyers (a visitor level variable is stored in a cookie, and based on its value you can change the website behavior for the users).

Finally, we recommend you a video of Custom Variables seminar delivered by Google. It will get you a better insight into this aspect of Google Analytics.