11 replies to this topic

[Vertster]

So I am sitting here looking at a site on clicktracks, and in the last month, there are approximately 13,000 search engine visits. Of these, 7000 of them have an average session length of 1 second.

First of all, I thought clicktracks filtered out bots, am I crazy to think this?

Who are these people? I am very familiar with the general click happiness that exists online, but 60% of the visitors are bouncing off like super balls!

[mcanerin]

There are at least 3 possibilities. Since I haven't seen the site in question and it would be good for the answer to help others, I'll outline the 3 that come to mind without regard to your site specifically.

First, there may be a problem with the page - perhaps it uses Flash and your visitors haven't upgraded yet, or maybe it loads too slow. The slow loading one happened to me once - the splash page didn't load up on time for dialup on time and they left. That will often take more than a second, though.

Second, your site may be getting pulled into another site (stealing content). That would create a quick load but no further surfing. But you said it came from search engines, so I don't know how that would work.

Third, it's possible that the people who are doing the search are not looking for your site but it's coming up high on the lists anyway. Is it possible that while optimising for one keyword set you've managed to score well on another, unrelated one without knowing it?

An example would be a site ranking well for a 3 letter aconym that is used by 2 totally different sites - A search for "SEO" on Google right now returns "Sponsors for Educational Opportunity", for example. Sometimes people click on the first link in the results without even really looking at it, especially new users.

There may be other reasons, but those could be some of them. A close look at the logs would help tell you. What was the search criteria, and what page was it - the main one of a sub-page? Good Luck!

Ian

Edited by mcanerin, 28 August 2003 - 04:36 PM.

[Jill]

I don't know if ClickTracks filters bots or could know all of them, but it sure does sound like it could just be spiders of some sort.

BTW, Vertster, it was nice meeting you in San Jose!

J

[Matt B]

As far as I know, ClickTracks has one of the best bot filters on the market. I think Ian hit on a couple other viable possibilities. Other than that, it would take some in-depth investigation.

Maybe open your log file in NotePad, and try to find a couple entries with 1 second visits, then try to find some correlation among the them and the rest. Find out where they are coming from by referrer or source. Nothing beats reading an actual log file for a couple of hours . . .

[Vertster]

First, there may be a problem with the page - perhaps it uses Flash and your visitors haven't upgraded yet, or maybe it loads too slow.

Sometimes its the obvious.... This is a frustrating circumstance, because this company has an in-house web person, who knows a lot less than the president thinks! I have told him time and again to keep the page weight under control, but it looks like he has mushroomed it back to 120k.

A pretty interesting case study in "consultant finds problem with site design without looking at the site code." Oh well, at least the changes we have suggested have doubled sales, even at 120k Now, get that load time under control, and things could get really interesting

[Vertster]

I don't know if ClickTracks filters bots or could know all of them, but it sure does sound like it could just be spiders of some sort.

BTW, Vertster, it was nice meeting you in San Jose!

Yeah nice meeting you! I wish I had found ya before I was half in the bag though...

[lottie]

Does anyone know if there exist statistics on website visit duration and where I can find them?

[Jill]

Welcome, Lottie!

If your Web site has a decent log file analysis package, you can find out the duration that your visitors are on each page. Is that what you're looking for?

Jill

[lottie]

Thanks Jill

I have details of visit duration on our stats package, but a client of ours is asking how long people stay on a site on average? 42% of their visitors are leaving the site within 30 seconds. How does this percentage compare with other people's experiences?

[daniel]

Hi Lottie.

Don't know about any stats based on empirical research, but think about your own browsing experience.

You search for something on Google, and are presented with a list of results. Of course it depends on what you're looking for and on the quality of the results, but more than likely you'll start towards the top and click on the links to see if the sites have the information you were after.

If not, you'll leave each site and return to the search results and try another site or refine your search.

30 seconds sounds about right to determine if a site could be offering the information you're searching for. Usually, I'd expect it to be far less that 30 seconds, but it depends on the nature of the site and the search.

I know it's a bit anecdotal, but I hope it helps.

Cheers,

Daniel

[Matt B]

Lottie -

Be sure you are breaking out the actual site visitors and any visits from bots. The bot visits may really skew those numbers if they are not filtered out.

Otherwise, an exit rate is something that is hard to determine, as every site has a different dynamic. However, 42% seems high.

My guess is that you have some bot visits mixed in there, otherwise, does your site target a subject that may also be interpreted as something else? Check your referral phrases and make sure they are accurate to your target audience. If everything checks out - the next step is to evaluate your homepage.

[mcanerin]

One of the most important, and least appreciated, stats that a log file analyzer can give you is the exit page.

That's where the visitor either gave up on your site or found what he or she needed - it's one of the closest things you can get to a "customer satisfaction" statistic without actually having polls and other more complicated measurment tools.

Is the main exit page your splash page? That's a big problem. Take a look at it - does it have clear links to the inside that bots can follow? Does it load too slow? is it spammy? does it clearly tell visitors what the site is about?

If the order page is your main exit page, then you compare that to your actual sales - the difference is often a strong measurement of your ability to close the sale. Are your prices too high? Are you asking someone to "buy" without telling them the price? Is your shopping cart unclear?

If the main exit page is your thank you page, then congrats!

You can combine the exit page stats with most of the other stats to follow the visitor through your site, and you can pick up a LOT of great information!

Ian