bounce rate on event page

[dishu]

Hi all

I have a movie player page - when the user comes there and watches the movie (the movie takes 2-3 minutes usually) and then leaves the site - (which is often the case that he doesn't proceed anywhere else on the site) -

is he considered a bounce then??

thanks!

[jimbeetle]

Hey dishu,

You offered something, the user got what they wanted, who really cares if it might be a bounce; and in this case I don't really think it is.

Though bounce rate as a possible metric has been tossed around a lot recently, it's very problematic. Why did the visitor bounce? It could be they found all the information they needed, but wanted to see if there was any more out there; that doesn't mean that the first bounce was not relevant.

In your case with a movie player I might consider a line up of other relevant movies or the ubiquitous "Folks that viewed this video also viewed these videos.." type of thing. Anything that can keep them on the site a bit longer.

[beu]

I agree, if users watch the video I wouldn't consider it a bounce.

You might check "time on page" and see if the average is equal to or greater than at least the playtime of the video.

[dishu]

thanks!

the thing is that if I look at the bounce rate for a date range - it's 70% but when I go to Navigation Summary I see that 76% proceeded to next pages...

Does this make sense?? Should it be 30% moved to next pages if the bounce rate is 70%?

[beu]

What analytics are you using and how do they define a bounce?

[j0nyDzine]

Any idea what let's say *most* packages consider as a bounce? Or does this vary too much? (Am I setting myself up with my own answer ? )

[beu]

Originally Posted by j0nyDzine

Any idea what let's say *most* packages consider as a bounce? Or does this vary too much? (Am I setting myself up with my own answer ? )

Not at all, great question!

Because there are no real "standards" in the web analytics industry, terms often have different meanings depending which analytics suite you are referring. IE "Unique Visitor" in one suite may not be calculated the same as "Unique Visitor" in another and might actually be called something like "Absolute Unique Visitor" in a third. As you can imagine, this gets tricky when you have different clients all using different analytics.

[jewboy]

It is my understanding that a "bounce" is when one page only is viewed, not whether a desired action is taken on that particular page.

[NewKidOnTheBlock]

I'm interested as much in web analytics as I'm interested in SEO and web analytics professionals usually use bounce rate to refer to unique visitors who come to your website and leave 'instantly' (which they sometimes define differently (seemingly) depending on what they feel is most insightful, but usually it's 0-5 seconds (some tools might use 0-11 seconds)).

If that's the definition your analytics package uses then viewing a video of a few minutes and leaving would definitely not be considered a 'bounce'.

Another way of which I've heard that some tools measure bounce rate is 'people who only see one page on your website'.

I'm not sure if you do so dishu, but some people confuse bounce rate with exit rate. Exit rates for a whole site would be 100% as everybody exits some time and are measured on a page-level..and most web analysts agree that exit rate is usually a quite meaningless metric, because people have to exit on some page of your site, which doesn't mean that it's a bad page (except for one exception: a funnel process (exp?)..when people are supposed to make several steps in order to buy something, in that case a high exit rate for a page in the funnel is something that needs to be looked at obviously!).

Bounce rate really is supposed to tell you how many percent of your site's visitors came to your site, were disgusted/didnt find what they were looking for, left.

So if your tool measures the process of leaving immediately using time then watching a 2 min video and leaving won't increase your bounce rate. If they use 'only one page viewed' to define bounces, those people would only be considered bounces if the page with the video on it was the very first site they saw on your page.

Hope anything of that made sense? couldnt hold myself back from commenting on analytics

P.S.: For the full formula you would have to divide them by number of unique visitors overall, of course..didnt want to make it more complicated than it needs to be :-).

[dishu]

Hi NewKidOnTheBlock
Thanks for the reply!

Well, we are using Google Analytics and for them the definition is "Bounce rate is the percentage of single-page visits"... Hm.

Could you tell me more about "divide at unique visitors number"?

Thanks!!!

[NewKidOnTheBlock]

What I meant by that is that if you want bounce rate you have to take the number of unique visitors who "bounce" and then divide them by your overall number of unique visitors, so you have "bounce rate" not just "number of bounces".

Sorry if I confused you!

[dishu]

no-no, you didn't confuse me!)

we have the Bounce Rate figure itself in Google Analytics

[beu]

Originally Posted by dishu

Well, we are using Google Analytics and for them the definition is "Bounce rate is the percentage of single-page visits"

Yeah, that's what I was afraid of!

Since you have video you're bound to have a high "bounce rate" by definition.

You could create an image on one page that users would click to take them to another "video page" and disallow the "video page" via robots.txt.

Is it really worth all that effort and blocking your video from being seen by engines?

Couldn't you just ignore the bounce rate in this case?

[Jazajay]

Just to add clarification/confusion, and another way at looking at it.

I count my bounce rate as the amount of people who leave your site from any page regarless if they arrived at that page internally or externally.

To me this gives far superior results.
That way you see which pages you have a problem with overall not just how many people clicked from the SERP's. You can then create a nice table in your CMS that has nice arrows red for a increase in bounce rate for that page, green for a decrease in the bounce rate and grey for the same/no change.

If you are bored this is a boring way to spend your day, but it looks pretty afterwards and give you an over view of your whole site and which pages are perfomaing and which are not. You could then do a nice list of top 10 under profming pages and top ten best prefomance pages but thats just me

You can then split it up to your -
bounce rate overall per page,
bounce rate from external sites/off the SERP's
bounce rate from internal pages.

I would then split the bounce rate futher.
External and SERP's
-External sites and which ones.
See if a certain link has a high bounce rate then either ask for the link to be reomved so as not to mess with my pretty table or to change the page/anchor text on the external site.

-SERP's
Then you know it's a problem with your click though rates and people not finding the information you are promising in the SERP description. In which case you need to change your SERP description to lower your bounce rate and increase sales.

Internal
Which pages where followed to the page and see if there is something missing. Bad internal linking that is making people think I have some thing when I dont or just bad information arcutecture full stop.

Dell lowered 1 of thiers by noticing people where expecting more information on 1 page. They added a calculater, to calculate finance or something, and as a result dropped thier bounce rate down to 35% from 78ish.

I agree, if users watch the video I wouldn't consider it a bounce.

I would disagree buts that because I'm arkward and backward human being.

Personally any time a person leaves your site from not completing the macro action, your over all goal - be it a sale, adsense click through or contact, it is a bounce as you failed to convert them to a customer. They bounced out of your capatist hands if you like.

it's 70% but when I go to Navigation Summary I see that 76% proceeded to next pages...

Thats possible if the bounce rate is calculated from only your site vistors hitting your sites from an external source, SERP's, external site.

The 76% proceeded would them be calcualted from internal hits to that page and then them carrying on.
Does that make sense?

I cant see how it could be calculated any other way to be honest. Why it would be caluclated like that god knows

In your case with a movie player I might consider a line up of other relevant movies or the ubiquitous "Folks that viewed this video also viewed these videos.."

That is spot on advice to lower it in this instance. The related is by far the better option if done right.

The best way to do it would be to write a computer algorithum to do this for you. There are books that you can buy to help you accomplish this.

The other way would be to manually add 5-7 related videos manually to a database that you think are related. Then dynamic populate those links to the page with the correct video.

Failling that the other people viewed would help to lower it but by no means as much as the related video option. Mainly as no one is veiwing any more of your videos.

How I would do it?
I would provide a menu to the left of the video with a h2 heading saying related videos with a border and maybe a differnet background colour. List them via database output or the computer algortihum. Then on hover provide a AJAX call and lift a thumb nail of the video on to a visable pop-up div. Place a nice call to action like watch this video button (make it large just below the thumb nail). You could put it in the DB and populate it via none obtusive JavaScript I suppose.

Then provide about 20-40px padding with the video to it's right.

Depending on your site I may make changes to that but that seems the best bet.

Why so precise?
Well by providing it to the left of the video you are making people scan over it to get to the video. If it was to the right they wouldn't see it due to tunnel vision for the video/information they want.

The header would then draw the eye and the seperate background would make it more distinghed from the rest of your content. The 5 to 7 links would just be enough any more and people would ignore it, too much information, less is more.
I unfortunatly haven't learnt that yet.

The AJAX thumb nail would then be a nice call to action to click through/ persavive actutecture/user experince.

You will then, hopefully true testing would be needed though, have more people clicking through to the rest of the site and completing a micro action, a step towards the macro action, the sale or whatever.

That should, if not will, lower your bounce rate no matter what you use to work it out or your definaition of the term.

Not bad from a self taught genius and off the cuff as I wrote it.
You don't need to say your good as I already know. It would be nice though -

The workings of a mad, yet loveable, English man.
Jaza