PDA

View Full Version : Google Analytics



orion_joel
02-03-2009, 08:36 PM
Sorry for so many posts here lately, but lots of questions as always.

Does anyone know what the delay on Google analytic's is (eg from user visiting site, to that visit being shown in the statistics)?.

nealrm
02-03-2009, 10:08 PM
It's around 24 hours.

jem
02-03-2009, 10:27 PM
I've actually never measured it - got me thinking though, I'll keep an eye on it!

Simon

orion_joel
02-03-2009, 11:04 PM
I suppose the hardest thing is actually knowing when the person actually accesses the page, and then being able to relate that specific visit back to when it shows up on analytics. I am sure that there must be a way it is just maybe it is not worth the trouble to work it out.

vangogh
02-04-2009, 12:46 AM
I don't think it's as long as 24 hours. I think it's only a few hours, though I've honestly never checked. Sometimes during the day I'll login and change the dates they're showing to include the current date and I'll see numbers for today. I probably never do that though till later in the day so it would have been at least 12 hours. I'm pretty sure the numbers will show up sooner, but I guess I've never checked earlier.

nealrm
02-04-2009, 07:34 AM
I'm surprised, the results are updated much faster than I thought. The results do trickle in throughout the day, but you wont get the full results for today until tomorrow.

The screen shot below was taken at 6:24 am and it shows traffic for the 5-6 am period.

http://farm4.static.flickr.com/3414/3252391655_3fcaf41f89.jpg

here a link to a larger photo, it's easier to read. Click here (http://farm4.static.flickr.com/3414/3252391655_69ca627cf5_o.jpg)

vangogh
02-04-2009, 10:01 AM
I guess that's proof. Neal you're right that you aren't seeing all results until the next day, but I think what you're seeing in the image is about right. I've never tested it myself, but I'd always heard it took about 4-6 hours to starting seeing your traffic in Analytics.

phanio
02-04-2009, 10:09 AM
If these numbers were updated instantly, what could/would you do with the information?

nealrm
02-04-2009, 02:04 PM
I just rechecked the numbers from this morning. It appears the update time is over an hours. The number of visitors reported for the 5-6 am time period increased since my last post.

rezzy
02-04-2009, 02:12 PM
If these numbers were updated instantly, what could/would you do with the information?

I think the real benefit is knowing if something happens and suddenly your stats spike, you can take steps to mitigate the damage. Outside of being able to see them quickly, I do not see alot of benefit.

Outside of Google Analytics, there are quite a few tracking systems.

For instance, ClickHeat, maps out where users are clicking on your site. It shows an actual page and where people are clicking on it. This runs from your server*

Woopra.com is another one. Each of these has its own pros and cons. Woopra runs using a js file like Google does.

orion_joel
02-04-2009, 11:15 PM
What more concerns me is how to link multiple known facts together. For example i post a link on twitter, using tweetburner, and it shows 10 clicks. I would assume that this is determined by people clicking on the link in their twitter feed. Aw Stat's showed at least a 10 unique user increase in this time. Yet Google Analytics shows exactly 2 visitors for this day, maybe it is still being finalized i do not know but, how the numbers can vary so much is beyond me.

rezzy
02-04-2009, 11:21 PM
I think how the stats are gathered are different.

And can included different items. Spiders may be included in Awstats.

orion_joel
02-04-2009, 11:42 PM
That i understand, yet how would the stats on tweetburner, and awstats roughly match, yet analytics is that far off. Also i am basing this theory on spiders not picking up the links in tweets? Would this be correct or not?

vangogh
02-05-2009, 10:25 AM
A lot of it can come from the difference between using javascript to capture the data and using the server logs. The javascript way shouldn't be collecting data from automated traffic, where the server logos will.

orion_joel
02-05-2009, 07:52 PM
Actually another thought i just had, for java script does a user need java installed on the computer, or will java script execute itself???

billbenson
02-06-2009, 03:03 AM
Java and JavaScript are two different things. If security in win XP is set to "high" JavaScript is disabled. Also, somewhere in the windows menu's there is a check box for disabling JavaScript. I forget exactly where. How Vista handles it, I have no idea.

Java is a server side language. JavaScript is client side. Beyond that, I don't know much about Java other than its a Sun Microsystems product. I'm not sure if JavaScript is a Sun product or not?

vangogh
02-06-2009, 10:33 AM
JavaScript isn't a Sun product. The only reason for Java in the name was an attempt at piggybacking off the success of Java. The two languages have nothing to do with each other.

orion_joel
02-06-2009, 07:51 PM
But potentially if the settings are right on a users computer javascript may be ignored?

billbenson
02-06-2009, 11:40 PM
I do everything like form error checking server side, mostly because I know how to do it in php and don't know js very well. The only js I use in most cases is a contact us link and that is to hide it from email spiders.

The direct answer to your question is yes the site visitor can turn off the ability to execute js on his pc. High security settings in XP will do this or they can do it manually.

I talk to a lot of customers daily and find it rare that my contact us link can't be seen which means js is enabled. You couldn't read half of the yahoo page if it was disabled. I really doubt it is an issue.

When doing form checking, do it server side or both. I wouldn't recommend just doing that in js. That's a security issue.

orion_joel
02-06-2009, 11:53 PM
Thanks for the response Bill, more the reason that i was asking the question was as a result of the differences between my awstats unique visitor count, tweetburner referred traffic, and analytics.

I suppose whatever the stats say if i am tracking them from a consistent source, then i should be at least getting relatively consistent figures for increases or decreases.

billbenson
02-07-2009, 12:09 AM
You could toss adsense at the bottom of the page to see if the hits match? they are js. Probably a js hit counter out there as well, which would just be one line of code. I wrote one a number of years ago, but I don't know if I still have it.

orion_joel
02-07-2009, 12:42 AM
For the time being i may just stick to Awstats, even though it takes in the spiders as well i know it is going to catch everything and hopefully will be a good representative view of changes to traffic.