Why you shouldn’t blindly trust Google Analytics reports

Want to unlock the full potential of your Google Analytics reports? Whether you’re an experienced digital marketer or just starting out in web analytics, follow our tips to get a truly deep understanding of the performance of your digital channels and digital campaigns.

Subscribe to our blog

When starting with site analysis to gain insights on how visitors are using your website, Google Analytics (GA) is by far the most popular tool of choice for digital marketers and web analysts around the world, including myself.

The reason for its popularity is quite obvious. Google Analytics is free, easy to implement and provides lots of data to use for improving your website in the form of default reports. While this is all very useful, some of the numbers found in GA reports are often wrongly interpreted or, in essence, inaccurate.

Time on page and session duration

To fully understand this metric, we should go back a little and explain how Google Analytics reports works. You can add tracking to certain, or all, pages on your websites. When someone visits one of these pages, Javascript code is executed when the page is loaded and collects information about the visitor's browsing behavior.

This has a major impact on two metrics: average time on page and average session duration. Since this JavaScript is executed when loading a page, time on page is tracked by calculating the time between two page loads.

For example: a visitor lands on the homepage of a company website, where the JavaScript is executed with the timestamp being 09:15 am. Then, he clicks on the products page, where the JavaScript returns a timestamp of 09:17 am. A simple calculation by Google Analytics would learn that the visitor spent 2 minutes on the homepage.

While this seems like a perfectly calculated metric, the catch is when this visitor exits the website from the products page. Since no new timestamp is created after this, Analytics has no idea how much time this person actually spent on this page. Might've been 10 seconds, might've been 1 hour.

Page

Homepage

Product Page

External page

Timestamp

09:15 am

09:17 am

??

Time on page

2

??

??

This means that a lot of pages aren't included in the  average session duration, which is an average of the sum of all sessions.

Average time on page is only a good indication when a page does not have a high Exit rate. High exit rates mean that this is the last page for a lot of visitors, which means a lot of unrecorded time that they might have spent on your website.

To counter this lack of data, you can configure and track custom event tracking through Google Tag Manager. You can define and measure predefined actions on a page, such as clicking a "more info" button that sends the visitor further down your page, or a “download” CTA button that prompts the visitor to fill in a form. That way, Analytics gets an extra timestamp.

Bounce rate

The bounce rate is the percentage of visitors that leave after seeing only one page of a website. These single-page sessions are seen as something negative, thus having a high bounce rate makes digital marketers think there's something wrong with their website because visitors don't interact with multiple pages. But is it really the case?

A one-page session might indicate that the visitor found exactly what he was looking for on this one web page. Because of the way Analytics works (cfr. " Time on page and session duration") a visitor might have been reading this one specific page in depth for 30 minutes but is still regarded a "bounce".

Or what if a lot of visitors are landing on your contact information and contact you by e-mailing or calling? This is an extremely valuable prospect, yet this session contributes to an increase of the bounce rate. Even filling out a form might not be registered by Analytics in certain cases, while this is the ultimate conversion goal for a lot of websites (e.g. to download content, request a sales quote or register to an event).

As a result, the default bounce rate metric is completely useless for single-page websites, since it is not even possible to view more than one page in this case. When analyzing a single-page website, custom tracking and metrics will need to be set up in order to get relevant results.

So before worrying too much about a high bounce rate, determine whether a bounce rate is relevant for your case and try to focus on certain web pages instead of looking at the average bounce rate of the complete website.

Direct traffic

In essence, direct traffic is a metric that shows how many people navigated directly to your website, e.g. by typing the URL in their browser or clicking on one of their bookmarks/favorites.

What could be, in essence, a valuable metric, is contaminated by Google itself. Analytics  attributes every visit it doesn't know the source of to the "direct traffic". This can include links in Microsoft Word documents, slide decks, PDFs, untagged email campaigns, etc., which makes it a bit “the garbage bin” of the source traffic metrics.

Instead of acccounting for "direct traffic" as visitors who already bookmarked your website, look at this data as visits without a traceable traffic source.

Conclusion

Whether you’re an experienced web analyst, digital marketer, or just starting out in web analytics, the default Google Analytics reports in are an extremely valuable source of insight into your website traffic and visitors’ behavior. But you should always maintain a critical mindset in order to filter out dubious data and ensure a sensible use of these metrics when you’re assessing the monthly performance of your digital channels and campaigns.

In order to unlock the full potential of Google Analytics reports, it is often better to use custom metrics. They will help you produce more flexible reports, that best describe your data and also provide a clearer context to the standard metrics.

Taking as an example the out-of-the-box bounce rate which we discussed previously, it might help provide some context to analyze the “non-bouncers”, a custom metric to measure all the people that don't bounce (leave after the first page). You can then drill down further into their behaviour with more refined metrics, and find, for example, what interactions made them continue browsing. Better analysis capabilities will result in a deeper understanding of your online audience, to support your content optimization and digital marketing activities.

Published on    Last updated on 02/07/2018

#Digital Marketing

About the author

Sam Vanhemelryck is Business Consultant Web Projects at Amplexor, based in Belgium. Sam is Google & Hubspot Certified and specializes in SEO, SEA, Google Analytics, and user experience and conversion optimization.

SUBSCRIBE TO OUR BLOG