Analytics

[AlexManifestMan]

We want to upgrade our analytics but man, the really really great ones seem to be at least $1100 month for all the features we want. We currently track every image and adjust them for conversions, track paths etc as well as track exactly where people are clicking on every page.

Any suggestions?

[gaydemon]

They do seem to be very expensive. I used to have a program called "Clicktrack Optimizer" which was pretty good except very slow.

In the end i found Google Analytics to be just as good and free. I must say Google Analytics (http://www.google.com/analytics/) is perfect, it can do far more than most people think.

Have you tried it?

[AlexManifestMan]

Yes, but one of the most important things for me is navigation paths and individual image popularity and I can't see how to set Google up to do that. We do use it (always two types of tracking to cross reference).

[gaydemon]

There is actually a navigation path or overlay i think they call it, which shows how people navigate or go. Not sure how good it is, ive always foudn those overlays difficult to make sense off due to my own sites setup.

[AlexManifestMan]

One of the ones we are looking at actually uses heat mapping to determine exactly where surfers click on each page and the number of times (to best and most quickly determine the popularity of particular images and balancing their effectiveness.

[gaydemon]

wow! That does sound very advanced. It sounds like a perfect tool to have. I bet its a high price to pay though, right?

[rawTOP]

Google Analytics has heat maps... Go to Content Overview and click on "Site Overlay" under "Click Patterns". You'll see stats on every link on your site... Showing everything from number of clicks to the revenue generated...

Alex - Look into Google Analytics "events". Basically you can register an artificial page view. So if they click from page A to page B by clicking on image 123, you can put an onClick Javascript event on the link and then your funnel analysis will have the image as part of the path that they took to convert (or drop out of the funnel)... Google "urchinTracker" and "pageTracker._trackPageview" to learn more...

The limitation with Google Analytics is that they only let you set up 4 goals. But if you use the e-commerce portion you can go way beyond that and track every product you have and break them down by category.

I've only set up goals, not e-commerce, but if you want a Google Analytics account with a fair amount of data in it to look at and play around with (to see if GA will do what you want), let me know and I'll permission you for reports for my site...

[AlexManifestMan]

That would be fantastic. I am really eager to try that. I believe that there is a clear relationship between images in the preview (some work better than others) but I want to really fine tune them. We are getting so many page views and I want to make the very most of it. Thanks SO much for that information.

A

PS. How do you reconcile the big difference in the numbers that GA gives vs something like AWstats. For instance AWStats shows 7141 unique visitors yesterday. But GA only shows 2836? That has confused me a lot.

[rawTOP]

Alex - send me a private message with the e-mail address you use to login to Google and I'll set you up to see my Google Analytics reports... If you already have Google Analytics set up my stats will only be slightly informative. I only do event tagging on outbound links to affiliate sites so I can get a rough estimate of conversion ratios by traffic source.

I don't see as big a difference between AWStats and Google Analytics as you're seeing... AWStats is about 10% higher for me - that amount doesn't surprise me too much. I'd guess you have spiders crawling your site that give user agent strings that make them look like regular web surfers, but because they don't execute GA's Javascript they get counted by AWStats (which is doing log file analysis), but not GA. I'd trust GA to be closer to the real number of genuine users. However, I'm working on implementing Unica NetInsight for a corporate client and we're seeing huge variations off Google Analytics there. Sometimes it's within 10% but a lot of the time Unica shows 50% or even 100+% more...

[gaydemon]

I have a 20% differences between my server stats and google analytics. Google definitly isnt 100% accurate, but its as reliable as you can get without analyzing your server logs. Plus ive found analytics to be so feature rich that i just stopped using my log analyzer.






PS. How do you reconcile the big difference in the numbers that GA gives vs something like AWstats. For instance AWStats shows 7141 unique visitors yesterday. But GA only shows 2836? That has confused me a lot.