Removing pages with high bounce rate; Will it improve my site's reputation?

[roodle]

Hi,

I manage a site that has a couple of sections with generally high (~80%+) average bounce rates. These sections of the site are not core material. One of the sections was being expanded up until quite recently. The other has been pretty much static for several years now. They both have a few popular entrance pages but visitors just tend not to go beyond the landing page. I do realize this could be seen as a positive also, but our intention was/is to draw visitors into the core pages.

Would the site benefit in terms of reputation/performance if I completely remove these pages from Google's index, so as to reduce the average bounce rate of the whole site?

Any opinions/experiences would be welcome.

Regards.

[Planet13]

I would look at using events to try and get a better grasp of some of the metrics. Like try to find out what percentage of visitors scroll down to the end of the main content area, or what percentage scrolls to the bottom of the page.

[Hoople]

Pages that provide everything that the user was looking for and doesn't require them to click outbound are normally high on bounce rate.

What was the time on page? Ditto to scrolling - did they?

Were all the keywords in their search found on the page as a part of useful to human info?

[roodle]

Hi,

Thanks for your replies. I will certainly analyze these pages a little more and add some events to both internal and outgoing links.

Of the typically 10-20% of visitors that do go into the site after landing on these sections, the average time on page isn't bad - over 3 minutes on some pages, which is encouraging.

Unfortunately most of the keywords used by visitors are not available via Analytics, but of those I can access it does seem like the information on page matches the queries.

I presume that to add scrolling events I'd need some special JavaScript - is there a piece of code you can recommend?

Also, going back to my original question - in principle, could removing these pages be a benefit to the site at all or would it be minimal at best?

[netmeg]

In my experience, pretty minimal, but it depends on whether you can figure out why people are bouncing off that page.

As mentioned above, if someone is getting the information they need and going off to continue their day, that's one thing. But if they're going back to Google because they didn't find what they wanted on your page, that's another.

I have a VERY high bounce rate on some of my pages, because a lot of people just come to find the date and/or time of an event and then take off. But I'm pretty sure the site is still considered to be high quality.

Do your pages consist of answers to the questions your users are likely asking? If so, it's probably fine.

[roodle]

if they're going back to Google because they didn't find what they wanted on your page

How is it possible to know if this is happening?

[Zivush]

Check: time on page for each page on the section you are specifying.
5 factors to take into consideration -

1. You'd want to explore what is the actual ranking of those pages in WMT (Google web master tools).

2. You'd want to explore what is the direct traffic from Google and the indirect traffic as well (how do people reach these pages?).

3. Is there a way to improve these pages or alternatively how they are presented to the readers?

4. Why do people bounce? Do they get what they were looking for.

5. Objectively/Subjectively - Are these pages efficient to the site/readers?

[Planet13]

I presume that to add scrolling events I'd need some special JavaScript - is there a piece of code you can recommend?

I would go to the analytics forum here on webmaster world and see if some of the analytics gurus there might have some better code than what I have been using.

I actually took event tracking OFF my site. Not because it didn't work, but because it worked TOO WELL!

After having it on for about 6 months, I realized that yes, my visitors WERE coming to a page, and the overwhelming majority were making it down to the bottom of the content area (like 70 percent or so), and about 15% would make it to the bottom of the page (down below all the footers / disclosures).

The thing is - at the end of the day - I really wanted to improve the bounce rate as measured by google analytics by default. And when you use those events, it was hard (for me, at least) to implement it without messing up the bounce rate.

So to summarize, I used event tracking just long enough to make sure that people really liked the content, and once I was certain of that, I removed the events.

Hope this helps.

[roodle]

I used event tracking just long enough to make sure that people really liked the content

Ok but surely scrolling doesn't actually imply that users definitely like the content. It usually takes time to read content to then decide whether it's not what you want. What other metrics did you use?

[levo]

The thing is - at the end of the day - I really wanted to improve the bounce rate as measured by google analytics by default. And when you use those events, it was hard (for me, at least) to implement it without messing up the bounce rate.

You can mark your event as "non-interaction."

[analytics.blogspot.com.tr...]

[Broadway]

Google the term "adjusted bounce rate." (I first saw it in a SearchEngineWatch article.)

It's where through Google Analytics code you add an event that gets triggered X seconds after the visitor lands on your page.

It's a way of estimating page engagement.

I have mine set to 60 seconds. Visitors don't register as a bounce if they stay on a page for over a minute.

It's a poor-man's dwell-time measurement.

[roodle]

Google the term "adjusted bounce rate." (I first saw it in a SearchEngineWatch article.)

It's where through Google Analytics code you add an event that gets triggered X seconds after the visitor lands on your page.

This is great information - thank you. I'd never heard of that feature.

Thanks to everyone for your input. I'll be re-assessing these sections in the light of your comments and suggestions.

[Zivush]

@roodle
There was a discussion long time ago called - "Is Panda all about Exit Rate?" [webmasterworld.com...]
I think you will find some of what was written there interesting and highly relevant to your current question.

[netmeg]

How is it possible to know if this is happening?

Well you have to kind of know your audience, but also, to the extent that you are able, look at the search strings. Are they queries that can be reasonably satisfied by looking at your page for 30 seconds or a minute, without going anywhere else? (Obviously you can't see (not provided) but you can maybe tell from other search engines, or your own site search if you have it)