Why You Shouldn’t Rely On Bounce Rate & Time On Site Alone

Feb. 27, 2012 | by Modestos Siotos

Analysing data is paramount in the day-to-day work of internet marketers so strategic decisions can be made based on scientific evidence. However, in some occasions some commonly used terms are sometimes misunderstood and used in the wrong context. Some of the most typical ones include:

  • Bounce rate
  • Average time on page
  • Average time on site

By and large, a high bounce rate is considered as a negative signal, often flagging the  need foe some conversion optimisation & usability improvements. Similarly,  low average time on page/site  is also considered as positive signals although these may not always be valid assumptions. An in-depth understanding of what these metrics represent is absolutely necessary, otherwise decisions may be made based on misconceptions.

The definitions of all major analytics metrics need to be well-understood before analysing the data of the various analytics packages.  For instance, the definitions for Google Analytic metrics s can be found here.

Below, we will discuss the flaws of bounce rate and time on site/page, and then define a better alternative, commonly known as dwell time.

Bounce Rate Flaws

This is the percentage of single page visits. However, bounces do not take into account the average time a user spends on a landing page. That means that when a user lands on a page from a search engine and stays just for a few seconds before exiting the site, it counts as a bounce. Similarly, when a user lands on a page and stays on that for several minutes before exiting the site, that also counts as a bounce.

In the first example,  the user did not find the landing page useful and left, whereas in the second example the landing page satisfied the user’s requirement.

So far, there are two takeaways:

  • Bounce rates should not be seen in isolation but in conjunction with other metrics (e.g. average time on page).
  • Bounces should not always raise concerns, as in certain cases a bounce can be a positive signal. This is when there is evidence that the user digested the content of the landing page before they happily left the site.

Time On Page/Site Flaws

This is another commonly used metric that also comes with some flaws. Time on page is calculated based on timestamps. Each time a user visits a new page on the site, the timestamp of the previous page is being calculated. That means that the average time on page of the last page before the user exits the site cannot be calculated and is always reported as 00:00. In other words, time spent on exit pages cannot be calculated. According to Google When a page is the last page in a session, there is no way to calculate the time spent on it because there is no subsequent pageview.”

As an example, if a user visits a page and bounces off, no matter how long they stay on the page, the time on page will be reported as 00:00. Consequently, reported figures for time on page can be skewed resulting in time on page/site values way off reality.

Dwell Time

According to Bing’s Duane Forrester, search engines can distinguish when the content of a page meets the searcher’s needs by looking at what he defines as dwell time.

“The time between when a user clicks on our search result and when they come back from your website tells a potential story.  A minute or two is good as it can easily indicate the visitor consumed your content.  Less than a couple of seconds can be viewed as a poor result.  And while that’s not the only factor we review when helping to determine quality, it’s a signal we watch.”

According to Dr Pete, dwell time is a combination of bounce rate and time on site. It measures how long it took a user to return to the search engine after clicking on a result. Unsurprisngly, there seems to be a strong correlation between dwell time and engagement.

How To Measure Dwell Time

In natural search, pages with low dwell time should be reviewed, analysed and improved so they offer a better user experience  . In a PPC campaign, low dwell time would increase the costs, flagging the page up as not relevant or of low quality. Although, analytics packages such as Google Analytics do not provide dwell time data, eye tracking technology can track it and it is one way to measure it (e.g. CrazyEgg, ClickTale).

 

 

Be Sociable, Share!

    Comments (9)

    • First Thoughts On Google’s New Tabbed Sitelinks | The best SEO Articale on the web

      [...] Arguably, the bounce rate on pages appearing in sitelinks will increase but at the same time sitelinks should drive higher quality traffic, potentially increasing user engagement and conversion as more users will be able to quickly land on the pages they desire. However, this will shift a great portion of traffic from the homepage to other pages on a site, essentially breaking the standard user journey into many shorter ones. The number of one page visits will also have a negative impact on the average time on page/site. When a user visits just one page and then leaves the site, the time on page will be shown as zero, which highlights the need to pay more attention to dwell time. [...]Jul 22, 2012 04:01 pm

    • Search Engine Optimization » Blog Archive » First Thoughts On Google’s New Tabbed Sitelinks

      [...] Arguably, the bounce rate on pages appearing in sitelinks will increase but at the same time sitelinks should drive higher quality traffic, potentially increasing user engagement and conversion as more users will be able to quickly land on the pages they desire. However, this will shift a great portion of traffic from the homepage to other pages on a site, essentially breaking the standard user journey into many shorter ones. The number of one page visits will also have a negative impact on the average time on page/site. When a user visits just one page and then leaves the site, the time on page will be shown as zero, which highlights the need to pay more attention to dwell time. [...]Jul 17, 2012 03:58 am

    • First Thoughts On Google’s New Tabbed Sitelinks | Lively Business Solutions

      [...] Arguably, the bounce rate on pages appearing in sitelinks will increase but at the same time sitelinks should drive higher quality traffic, potentially increasing user engagement and conversion as more users will be able to quickly land on the pages they desire. However, this will shift a great portion of traffic from the homepage to other pages on a site, essentially breaking the standard user journey into many shorter ones. The number of one page visits will also have a negative impact on the average time on page/site. When a user visits just one page and then leaves the site, the time on page will be shown as zero, which highlights the need to pay more attention to dwell time. [...]Jul 16, 2012 01:08 am

    • First Thoughts On Google’s New Tabbed Sitelinks | Ranking Maine

      [...] Arguably, the bounce rate on pages appearing in sitelinks will increase but at the same time sitelinks should drive higher quality traffic, potentially increasing user engagement and conversion as more users will be able to quickly land on the pages they desire. However, this will shift a great portion of traffic from the homepage to other pages on a site, essentially breaking the standard user journey into many shorter ones. The number of one page visits will also have a negative impact on the average time on page/site. When a user visits just one page and then leaves the site, the time on page will be shown as zero, which highlights the need to pay more attention to dwell time. [...]Jul 14, 2012 12:32 pm

    • First Thoughts On Google’s New Tabbed Sitelinks « Money Rail

      [...] Arguably, the bounce rate on pages appearing in sitelinks will increase but at the same time sitelinks should drive higher quality traffic, potentially increasing user engagement and conversion as more users will be able to quickly land on the pages they desire. However, this will shift a great portion of traffic from the homepage to other pages on a site, essentially breaking the standard user journey into many shorter ones. The number of one page visits will also have a negative impact on the average time on page/site. When a user visits just one page and then leaves the site, the time on page will be shown as zero, which highlights the need to pay more attention to dwell time. [...]Jul 13, 2012 10:00 pm

    • First Thoughts On Google’s New Tabbed Sitelinks | Money Rail

      [...] Arguably, the bounce rate on pages appearing in sitelinks will increase but at the same time sitelinks should drive higher quality traffic, potentially increasing user engagement and conversion as more users will be able to quickly land on the pages they desire. However, this will shift a great portion of traffic from the homepage to other pages on a site, essentially breaking the standard user journey into many shorter ones. The number of one page visits will also have a negative impact on the average time on page/site. When a user visits just one page and then leaves the site, the time on page will be shown as zero, which highlights the need to pay more attention to dwell time. [...]Jul 13, 2012 09:35 pm

    • Modestos Siotos

      Hi Mike,
      Even though a 90% bounce rate sounds like a very high percentage it always needs to be evaluated into context. The average time on page is a good indicator of user engagement with 3-5 minutes it sounds like your visitors are actually reading the posts. However, instead of leaving the site after reading a post there are ways to motivate them to stay longer (e.g. read a relevant blog post) or perform some other action by incorporating strong calls to action (e.g. subscribe to newsletter).Jun 29, 2012 09:07 am

    • Mike

      Hey Modestos,

      Great post. I was getting extremely concerned with one of my blogs. I was getting 90% bounce rates on quite a few posts but the average time on each page where between 3-5 mins.

      So from what you are saying then, that’s ok? Because people are reading the post before leaving/bouncing?

      ThanksJun 28, 2012 05:28 pm

    • Worried About Your Bounce Rate? — SEO Sugar Land TX « Local SEO Consultant | Online Reputation | Houston, TX | Rust Reviews, LLC, DBA IX Brand SEO Services Company

      [...] Why you shouldn’t rely on bounce rate and time on site alone (connect.icrossing.co.uk) 29.300652 -96.131493 Share this:EmailFacebookTwitterDiggStumbleUponLinkedInRedditTumblrPrintLike this:LikeBe the first to like this post. [...]Mar 31, 2012 10:20 pm

     
    Please note: the opinions expressed in this post represent the views of the individual, not necessarily those of iCrossing.

    Post a comment

    SUBSCRIBE