Google Analytics Bounce Rate Mystery- please explain!

[Ellen C Braun]

Hi,

Google Analytics is pretty straightforward, but regarding bounce rate - I cannot understand something:

Yesterday Google sent one of my sites over 200 visitors.

Every visitor who "bounced off my site" for a keyword, which showed a 100% bounce rate, stayed on my site for ZERO seconds. (This looks suspicious to me.)

Many visitors stayed on my site for much longer, but ALL OF THOSE VISITORS had a 0% bounce rate AND went to at least TWO PAGES OF MY SITE.

It is statistically impossible that out of 200+ visitors, nobody (of about 100 visitors) who visited just one page of my site stayed on it longer than zero seconds. What's up with that???

By the way, as an affiliate marketer, I am fine if people visit my site, read my info for 5-55 seconds, and then click from my site to Amazon. Does Google think that is bad??

It's quite possible (and VERY likely) than many people read my article for 30 seconds- 5 minutes, and then click through to my affiliate link. Google is actually giving me a complex by telling me that anyone who does not click through to another page within my site just stays on my site for ZERO seconds - that is totally false!

(As an affiliate marketer, I'm very happy if people click through to my SE listing and then to my affiliate link without reading other pages of my site... why does Google think it is BETTER if people read multiple pages of my site? They found the info they wanted from me, now they are clicking to buy!?)

And, on a related note, does Google penalize me via lower SE rankings for having a 100% bounce rate??

Thanks in advance

Ellen

[Bingo123]

I've also noticed a lot of traffic to my site says it was there for 00:00 seconds. Not quite sure what thats all about either?

[bgmacaw]

First of all, you're confusing 'bounce rate' and 'visit length'.

Bounce rate is the percentage of visitors who only visit one page on a site.

The visit length is the amount of time a visitor spends on a site, regardless of the number pages on the site visited.

Bounce rate and visit length depend a lot on the nature of the site. For example, an ecommerce site would want to retain visitors and have them move about the site. However, a site like Drudge Report or Digg would expect a high bounce rate and short visit times. This, coupled with typical browser user behaviors, would make these stats difficult for Google to use with any accuracy in ranking a site.

As for the specific traffic pattern you observed, this could be due to 'bots or some other automated process. For example, there are several popular IM programs that scrape Google search results and scrape/analyze the pages found in the results.

[Ellen C Braun]

Thanks, Bgmacaw, but I am NOT mixing up the "bounce rate" and the "visit length".

I am confused as to why EVERY SINGLE VISITOR who only goes to ONE page of my site (hence a 100% bounce rate) is recorded by Google as having stayed on my site for ZERO seconds???

Why, out of 245 visitors yesterday to this site, does Google show that THERE IS NOT ONE VISITOR WHO ONLY WENT TO ONE PAGE OF MY SITE, BUT WAS ON IT FOR MORE THAN ZERO SECONDS?

Can it be that every single visitor who only goes to one page of my site, also only stays on the site for zero seconds?? Statistically impossible.

All of the visitors who stayed on my site for any length of time, visited at least 2 pages of my site.

It seems that if a visitor only goes to one page of my site, google calculates their time on my site as ZERO SECONDS, even if the time was truly ten minutes!!

Who can solve this mystery?

[dotcomken]

Originally Posted by bgmacaw

As for the specific traffic pattern you observed, this could be due to 'bots or some other automated process. For example, there are several popular IM programs that scrape Google search results and scrape/analyze the pages found in the results.

Google Instant will not trigger google analytics tracking anymore as of a few weeks ago.

I do agree that the :00 time on page could be triggered from a bot or another automated program. Sometimes iFrames can even confuse the script.

What to look at is your avg time on page metric, Unless we can see who the actual visitor was beyond the hostname it is not worth our time.

There is also a latency of data in GA, sometimes it can take a few days for the data to be accurate. I have seen reports run a month apart and both for the same time period only to show different stats.

I will pm you a report I use to measure page efficiency. I'd link it but haven't posted enough on here yet...

..

If you want it please send me a DM @dotcomken on twitter and I'll get you the custom report to share.

[Ellen C Braun]

Do a google search for "ellen c braun is the greatest parenting coach in the universe"

and click through to whichever page on raisingsmallsouls.com comes up.

Stay on it for a full minute, without clicking to any other page, and then I'll check what my stats say for that keyword phrase!

[Dan B Rusu]

Originally Posted by Ellen C Braun

Do a google search for "ellen c braun is the greatest parenting coach in the universe"

and click through to whichever page on raisingsmallsouls.com comes up.

Stay on it for a full minute, without clicking to any other page, and then I'll check what my stats say for that keyword phrase!

Done. Just did that for you. Interested to see what happens...

[dotcomken]

Took a look at your site like you asked by using google.

A couple tips for your site:

1. Upgrade to the asynchronous version of the google analytics code.

2. There is also "var dom = {};dom.query = jQuery.noConflict(true);" within your footer.php possibly. Remove that. This isn’t going to work in WordPress, so at all times we use jQuery(document).ready(function($). The reason for this slight complication is that Prototype and Scriptaculous libraries arrived first into WordPress and have got to hold on to using the $ shortcut. And don’t be tempted to dabble with jQuery.noConflict() here – you won’t get good results.

3. Run your pages through a page load tester like in Chrome or a web app. Seemed to take a long time on some pages to load all of your scripting. This could be interfering with GA timers.

4. Lastly, have your developer install GAaddons and define your own bounce metric. you can also track page load times in google analytics.

[Ellen C Braun]

Originally Posted by Daniel7rusu

Done. Just did that for you. Interested to see what happens...

Thanks!

Ugh, multiple people did this, so it got messed up, I see 2 visits for this keyword, 3 pages per visit, avg time on site 1:19, bounce rate 50%!

[Ellen C Braun]

Originally Posted by dotcomken

Took a look at your site like you asked by using google.

A couple tips for your site:

1. Upgrade to the asynchronous version of the google analytics code.

2. There is also "var dom = {};dom.query = jQuery.noConflict(true);" within your footer.php possibly. Remove that. This isn’t going to work in WordPress, so at all times we use jQuery(document).ready(function($). The reason for this slight complication is that Prototype and Scriptaculous libraries arrived first into WordPress and have got to hold on to using the $ shortcut. And don’t be tempted to dabble with jQuery.noConflict() here – you won’t get good results.

3. Run your pages through a page load tester like in Chrome or a web app. Seemed to take a long time on some pages to load all of your scripting. This could be interfering with GA timers.

4. Lastly, have your developer install GAaddons and define your own bounce metric. you can also track page load times in google analytics.

This sounds like it's probably very good advice, but it is all completely GREEK to me! Uh oh!

[dotcomken]

Originally Posted by Ellen C Braun

This sounds like it's probably very good advice, but it is all completely GREEK to me! Uh oh!

The non-technical version of what I saying:

You have a lot running script-wise on your site. There could be something technically wrong with it that we do not see. Especially with the delayed load times do to external scripts. The javascript library you are using on your site (jquery) caught my eye. I know of the NoConflict bug, since I work on my own wordpress themes.

Look into optimizing you page load times and integrate the newer Google analytics tracking script into your site. Never know, could be a night and day change for you. Google does take load times into its quality metrics.

[Ellen C Braun]

Thanks- I am going to ask my tech guy to read this thread and see what he can do!

[patey88]

Originally Posted by Ellen C Braun

It's quite possible (and VERY likely) than many people read my article for 30 seconds- 5 minutes, and then click through to my affiliate link. Google is actually giving me a complex by telling me that anyone who does not click through to another page within my site just stays on my site for ZERO seconds - that is totally false!

It is normal for every bounce visit to be of zero length. I've always figured time spent on a page is calculated by subtracting arrival time on the first page from arrival time on the next page. Since the visitor never goes to a next page, you never know how long they spent on the landing page.

You can add "onclick" code to your outgoing affiliate links to make them be tracked as page views. If you do this, you can see visitors who click these links and then click "back" to your site, as if it's all within one visit.

-- Patey88

[Colema14]

Originally Posted by Ellen C Braun

Thanks!

Ugh, multiple people did this, so it got messed up, I see 2 visits for this keyword, 3 pages per visit, avg time on site 1:19, bounce rate 50%!

Parse your data by location/browser/OS or other ways to see which visitor is which. Its very unlikely that they all look the same.

This thread is old and I haven't read any further, but a previous poster mentioned instant previews and it just dawned on me that this is definitely the culprit of these type of visits. I read a great post in the google Analytics blog explaining it earlier this year, but since they view your page, but don't technically visit your site, I believe it's counted as a visit and a bounce with zero time on site. I'm sure someone else ends up solving this after this post, but I couldn't help but to chime in.

My problem is that I'm seeing this occur with my google Adwords and am being charged for it! I know that google just released this, but it doesn't seem fair that they count it as a click. Im going to see if they allow it to be disabled. It's happening to me across the display network though, so maybe I just need to opt out of that.

Zeek Coleman
Internet Marketing Manager of Visit Savannah
ZeekColeman.com