0:00 lenght visits what does that means to you

Explanation about 0:00 visits

[silverbytes]

I notice several 0:00 visits
I can understand that 0 to 10 seconds visits may be related to those who visit your website and expcted something different and leave.
But my worry is about those 0:00

What does exactly mean a 0:00 visit?
They are produced in different hours, different ips and countries. Some from normal search results, some from adwords, some other comes from other websites.

Some point to /index.htm but other to another pages.
Most of them recognized as Mozilla 4, 5 o 6 compatible.
And bytes out average 18 kb.

What are those 0:00?

[arran]

If your tracking software uses only the server logs to create these stat then that would explain it. A user who views one page might download your .html and .css in the same second then leave. In that case, the length of the visit would be 0:00 - the difference in time between the first and last server log entries.

arran.

[silverbytes]

Do you say they can leave before 1 sec?
Wow! My website must be a scarecrow!

[arran]

It doesn't mean they leave after zero seconds, it just means they only view one page. If there is only one entry in the server log (or several entries within the same second) and no entries thereafter, your tracking software has to report it as a 0.00 visit (even although the user could have spent 5 minutes reading your article). You don't generate any log entries when you hit the back button ;)

[joaquin112]

what about those annoying SPAM bots

[silverbytes]

It doesn't mean they leave after zero seconds, it just means they only view one page. If there is only one entry in the server log (or several entries within the same second) and no entries thereafter, your tracking software has to report it as a 0.00 visit (even although the user could have spent 5 minutes reading your article). You don't generate any log entries when you hit the back button ;)

Is that true? So if you spend 1 hour reading the homepage and click back on IE the software doesn't count the time spent and it's a 0:00 visit?

Any other possible cause?

[cgrantski]

That's right. If there's no 2nd page in a visit, there's no way to know how long they were on the 1st page. Logs track requests to your server, period. You can tell how long visitors were on a page by subtracting the time of the first page request from the time of the second page request. If there's no second page request there's no time measurement possible for the viewing of the first page. In the log, it's just a request for the page, then nothing more; the visitor disappears from the logs.

0:00 doesn't actually mean zero seconds, it means "length of visit unknown because there was only one page in the visit."

[silverbytes]

I really can't belive it (I do but I'm very surprised yet) why just don't count the history go -1 like another page (in fact the page were you was previously) and count that time since it's so common...

BTW how long do your visitors stay in your website. I discovered a 78% leaving before 30 secs.

[cgrantski]

I don't understand "why just don't count the history go -1 like another page (in fact the page were you was previously) and count that time since it's so common... " - can you explain a little more?

[silverbytes]

I mean I'm surprised that the click on back button on IE is not considered "another page" to be able to measure the time elapsed between the visit to your first page and, the click on back button.
History go -1 was reffered to the back action.

[cgrantski]

It's because what logs do is record requests (for files from your server). Clicks are, of course, requests, or rather they result in requests to servers. My point is that the request is what gets recorded in logs, not the click.

When using the back button while on your site, the request that results from the back button action is a request for a page on somebody else's site. So that request gets recorded in the logs belonging to the other site.

If you had the logs belonging to the other site, you could subtract the time of the request for your page from the time of the request for the other site's page (that was requested by the back button click) and you would then know how long they were on your page. But you'd have to have the other site's logs.

I'm oversimplifying of course. The back button click often doesn't get recorded in the logs of the other site either, because the browser usually has kept a cached copy of that previous page and instead of sending the request to the other site's server, the browser just pulls up the copy that it has kept on your server. That's just a detail and doesn't affect my main point which is that the back button click won't be recorded in the server's logs for YOUR site.

Hope this is clear. The difference between recording clicks and recording requests is a fundamental one in tracking that clears up a lot of misunderstandings. I was doing tracking analytics for about six months before I "got" it.

[maxi million]

As Arran says "It doesn't mean they leave after zero seconds, it just means they only view one page." The problem here is there is no record of where the visitor is going next. i believe most site analytics/stats programs you come across, calculates the length of stay on a page based on a simple formula:
length of stay on page 1 = (time of entry to page2 - time of entry to page 1)
correct me if im wrong here please...

now lets assume that the visitor has been on page 1 for a rather long period of time, but doesnt navigate to another page in your site or happens to follow a direct external link or simply closes the browser, there is record of where (s)he goes...

i like to think of it this way:
someone enters your house through the door and that door was opened by your butler who recorded the time of entry, but your visitor for whatever reason sneaks out through the window...
how can you know when (s)he left...?