« »
Apr 14
2009

A lot of time, the issue is not reporting nor analysis but it is making sure:

  • Data is accurate (or as accurate as it can get :) )
  • What we think we are analyzing is truly what we are analyzing

If you think one of your most important pages has a 100% bounce rate, you might panic (and rightfully so) and gather all your resident experts to figure out what is going on. Imagine, after hours or days of analysis and review you find out that all of your bounces come from a statistically insignificant number (in our example, 3) of pageviews! Most of us would not bother to investigate further and try not make the same false misperception again.

Let’s explore this issue in more detail.

Question: How does my e-commerce thank you page, which is not a landing page, have 100% bounce rate in the Top Content report in Google Analytics?

Answer: You are in the wrong report, my friend!

Let us start with the definition of Bounce rate:

  • “Bounce rate is the percentage of single-page visits in which the person left your site from the entrance (landing) page,” by Google Analytics.
  • “Single page view visits divided by entry pages,” by the Web Analytics Association.

So bounce rate is a metric for only landing pages and not for all pages. Reading the bounce rate for pages that are not entrance pages will lead to misperceptions, incorrect conclusions, and wrong actions.

What is the issue?

The issue starts when we try to make sense out of the bounce rate column in the content performance report. This report contains all pages visited in your site, including entry pages and non entry pages.

So let’s say we want to study and analyze the performance of the ‘confirmation.htm’ page and we started with the top content report.

When we look at the bounce rate on the content performance report, it is quite misleading since the 100% bounce rate is only applied to visits when this page was a landing page. One might ask, if this page is not a landing page and you have to go through a few steps before you reach it, where did these three visits in our above example come from?

There are a few scenarios where a non-landing page could be tracked in Google Analytics as an entry page (landing page). These scenarios include but are not limited to:

  • The thank you page was bookmarked by the visitor for future reference.
  • The visitor hit the refresh button on the non-landing page window after 30 minutes of no activity.
  • A direct visit to the non-landing page by developers/site owner. Excluding the internal traffic will solve this one.

If we look at the “Landing Pages” report, we can see these leaked pages and their bounce rate; 3 single pageview visits out of 3 entry pages leads to a 100% bounce rate.

Suggestions for web analysts:

I advice my dear analyst friends to not look at the bounce rate column in the “Top Content” report for non-landing pages. If you insist then I suggest excluding bounce visits with an Advanced Segment.

Now you have clean and accurate data

Suggestions for Google:

I suggest to the Google Analytics developers to remove the bounce rate column from the top content report all together. Or at least make it possible to remove the column from the display.

From now until the next blog post, I wish you a happy April and an enjoyable month of analysis. :)

Tags: ,

8 Responses to “Problems with the Bounce Rate in the Top Content Report”

  1. Rehan Asif says:

    I will add that subdomains or multiple domains without the right tracking code can lead to thank you pages being classified as entry/landing pages.

  2. Cassandra says:

    great post and I agree with you that there should be a way to remove the “bounce rate” column.

  3. Another thing to consider is that this can occur not only if the prior page has no tracking code installed, but if the form was a PDF file, or Flash form, both of which, can reside off-line and still post form data to the web server.

  4. Jeff Vogt says:

    I was getting a bounce rate on non-landing page and it turned out that this was the first page where I switch over to SSL. It appears Google counts https as a separate session.

  5. Rehan Asif says:

    Hi Jeff,

    Google Analytics is not supposed to count https as a separate session.
    Depending on how your site is configured and the transition to https from http, you may need to modify the Google Analytics tracking code that goes on every page.
    If you are not sure what to do to fix the problem, give us a call.

  6. how to fix it with ssl redirect says:

    We have a site that is full ssl so the domain goes like this:

    https://www.example.com

    in order to redirect all users who type http://www.example.com to https://www.example.com we are using and .htaccess rewrite rule.

    How can we fix the google code or do something in order to lower the bounce rate from users entering the site via http://www.example.com ?

    I noticed that all bounce is from / page

    Thanks for the help

  7. Rehan Asif says:

    If the Google Analytics tracking code (GATC) doesn’t load until they hit the https site there should not be any problems with URL rewrites or redirects.
    Can you determine if the GATC loads when someone requests the http site and before they are redirected?

    It is quite normal to have people bouncing from the homepage. What kind of bounce rates are you seeing?

  8. Lars Meyer says:

    Apart from flash, Ajax on your website can also be a problem when trying to measure your bounce rate correctly, as I have written about here: http://www.intramedia.dk/bounce-rate.aspx.
    Too many websites are not developed in a way that ensures the correct registration af Bounces with an ordinary Google Analytics integration. With complex websites and webdesign and can be a real pain to call/instanciate the analytics code manuelly og dynamically via javascript directly in the ajax or jquery calls.

Leave a Reply