Website Bounce Rate Measurements
I am interested in getting input on best practice approaches for measuring website bounce rate (and user experience).
We utilize Omniture as our website metrics tracking application. Within Omniture there is a standard definition of bounce rate that is defined by whether or not a user completes a second Omniture load event, which typically involves the user going to a second page. We have also implemented a second measurement that tells us the percentage of users who leave before spending five seconds on their first visited page.
Our goals are to:
-Determine if we are providing a strong user experience
-Understand the signals our site usage behavior sends to external sites, such as search engines
In many cases, a user can visit our site on a specific landing page, have a rewarding experience (in terms of consuming useful content) and then leave without visiting a second page. This counts as a bounce under some definitions of bounce rate (omniture's default), but not under others.
In short, I am wondering if anyone has input on bounce rate or user behavior measurements that they have used to identify the usefulness of a website landing page to a visiting user.
Thanks.
We utilize Omniture as our website metrics tracking application. Within Omniture there is a standard definition of bounce rate that is defined by whether or not a user completes a second Omniture load event, which typically involves the user going to a second page. We have also implemented a second measurement that tells us the percentage of users who leave before spending five seconds on their first visited page.
Our goals are to:
-Determine if we are providing a strong user experience
-Understand the signals our site usage behavior sends to external sites, such as search engines
In many cases, a user can visit our site on a specific landing page, have a rewarding experience (in terms of consuming useful content) and then leave without visiting a second page. This counts as a bounce under some definitions of bounce rate (omniture's default), but not under others.
In short, I am wondering if anyone has input on bounce rate or user behavior measurements that they have used to identify the usefulness of a website landing page to a visiting user.
Thanks.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
The current 'experts' probably never see your landing pages since we get here usually by clicking on a link in the notification emails or by looking at the results in our "Saved Searches".
One of the other problems with 'bounce' here is that if the user is looking for an answer to a question and they find it the first time they come here, then they will usually take the answer and go fix their problem. They might come back later because they found their answer but unless you can track that, you won't know that they had a "positive first experience".
I think you need a second action to actually get any useful information. For new unregistered users, you might put up a question and checkbox asking "Was this answer useful to you?" Not everyone will make use of it but it will be more informative than just the 'bounce' info.
One of the other problems with 'bounce' here is that if the user is looking for an answer to a question and they find it the first time they come here, then they will usually take the answer and go fix their problem. They might come back later because they found their answer but unless you can track that, you won't know that they had a "positive first experience".
I think you need a second action to actually get any useful information. For new unregistered users, you might put up a question and checkbox asking "Was this answer useful to you?" Not everyone will make use of it but it will be more informative than just the 'bounce' info.
This is a continuing discussion with my main customer. We get all these statistics and information but the one thing that is always lacking is 'why'. If we knew 'why' we'd know what to do. Until then, we just try to make our best guess and see what happens.
I'll just leave this here for the humor factor. ;-)
https://www.facebook.com/photo.php?fbid=10152518370439179&set=a.394263064178.194611.786539178&type=1&theater
https://www.facebook.com/photo.php?fbid=10152518370439179&set=a.394263064178.194611.786539178&type=1&theater
ASKER
Hi all,
As mentioned, there are two factors driving this request (and it relates to landing pages for new visitors to our site):
1. If a user leaves almost immediately after arriving to our site, it sends a negative signal back to the search engines that may adversely impact how the search engines rate our site.
2. If we see users leaving rapidly after arriving (ie, before they could have read any of the content on our landing page), it could imply that we are creating a negative user experience.
While bounce rate is a commonly used term, we have found that it is actually measured in many different ways (depending on context and on metrics package). Therefore, I am looking for insight on the ways that others have seen bounce rate measured in order to measure the two factors listed above.
To date, the best measure I have found is to set a time limit (ie, five seconds), and to measure what percentage of visitors have a time on site longer than this threshold. Our goal is to set a standard benchmark and then to test alternate page layouts and page treatments in order to find the page structure that results in the largest percentage of users staying on site long enough to consume our content.
As mentioned, there are two factors driving this request (and it relates to landing pages for new visitors to our site):
1. If a user leaves almost immediately after arriving to our site, it sends a negative signal back to the search engines that may adversely impact how the search engines rate our site.
2. If we see users leaving rapidly after arriving (ie, before they could have read any of the content on our landing page), it could imply that we are creating a negative user experience.
While bounce rate is a commonly used term, we have found that it is actually measured in many different ways (depending on context and on metrics package). Therefore, I am looking for insight on the ways that others have seen bounce rate measured in order to measure the two factors listed above.
To date, the best measure I have found is to set a time limit (ie, five seconds), and to measure what percentage of visitors have a time on site longer than this threshold. Our goal is to set a standard benchmark and then to test alternate page layouts and page treatments in order to find the page structure that results in the largest percentage of users staying on site long enough to consume our content.
Sounds like as good a plan as any. Where would these landing pages be found? As a qualified expert, I usually go directly to a page from the link in an email about the question. I probably never see the landing pages.
Ok, so in another browser where I was not logged in, I went to Google and found Q_22760371.html, First, it never seems to stop loading which bothers some people. The hour glass or rotating circle never goes away. Second, "Enjoy your unlocked premium solution" is an odd phrase. I know what you mean because I been here for years now. But I think that 10 years ago that would have caused a 'Huh?' response.
And I still think the Chicken Pizza advertisement is a little odd on a Tech site.
And I still think the Chicken Pizza advertisement is a little odd on a Tech site.
SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
Hi Ray,
I agree with your broader comments about User Experience - this is a topic area we need to study closely as we consider our long-term web and mobile experiences.
In the context of this question, our thinking is that we will use a five second bounce rate measurement in order to measure:
-The metric data we are passing back to search engines
-As a rough measurement of whether we are creating an acceptable vs unacceptable initial presentation of our content on our landing pages
Bounce rate will not tell us whether our broader user experience is acceptable, but it could inform us on which landing page approaches and layouts create a better first impression.
Barring other comment, we will assume that the 5 second bounce rate measurement is something around which there is general consensus (in regards to bounce rate metric configuration). While bounce rate can be a blunt measurement, it is a metric we want to use to its fullest potential.
Thanks,
Brian
I agree with your broader comments about User Experience - this is a topic area we need to study closely as we consider our long-term web and mobile experiences.
In the context of this question, our thinking is that we will use a five second bounce rate measurement in order to measure:
-The metric data we are passing back to search engines
-As a rough measurement of whether we are creating an acceptable vs unacceptable initial presentation of our content on our landing pages
Bounce rate will not tell us whether our broader user experience is acceptable, but it could inform us on which landing page approaches and layouts create a better first impression.
Barring other comment, we will assume that the 5 second bounce rate measurement is something around which there is general consensus (in regards to bounce rate metric configuration). While bounce rate can be a blunt measurement, it is a metric we want to use to its fullest potential.
Thanks,
Brian
Brian, it's a number, and I don't know whether 5 seconds is the right number, so I can't really say much constructively about it. I would just suggest that this is a great question to take to Google (and anything that EE can do to increase and enhance its dialog with Google is probably a good idea). We're still w-a-a-a-y off the mark in Google search results, and anything that helps bring EE results into the first page will be a big step forward.
Best,
Ray
Best,
Ray
ASKER
I am specifically referring to:
-The bounce rate in relation to a user's first page visited on our site
-The behavior around landing pages that users will access from search engine results, external links or by typing in a URL
We want to make sure that we have a clear measurement for whether the initial interaction to the site results in someone staying long enough to interact with the site's content (versus leaving immediately because they find the content irrelevant, have a bad user experience or a bad brand experience).
If we have the appropriate metric in place, we can test alternate page structures and alternate ways to present content and optimize our landing pages to make the first experience of a new user as positive as possible.
I am wondering if any experts have input on what the best measurement metric would be (whether it be one of the two we are using or an alternate approach).
Thanks
Brian