A while back I got curious about the kind of traffic I've been receiving on this blog as Blogger has nothing constructed to make it in anyway clear if your blog is getting traffic of any kind at all. So I decided to research a few different ways to do it and finally came to the title package.
Stat Counter as noted above was my the choice I eventually rested upon and I'm pleased I went with this solution. The system solved a unique problem I had with tracking this blog. Specifically the articles and resources for this blog, as well as it's delivery server are all remote systems that I simply submit content to. As such I can't add any sort of monitoring to the system or implement any form of logging (or even access the logs that are invariably generated). So I needed something that could be added directly to the pages code and would utilize an external source to make the magic happen.
Stat Counter pulls this off by inserting a small block of java script into your page. This does have a few failings, specifically in that I am still unable to track usage from the ultra paranoid (not that I disagree with their caution, even if i consider it somewhat unwarranted in this case). Stat Counter also provides a number of different ways this java script code can manifest on the page. I've chosen the simple button that will take you to the stat counter website but there's also a number of counter options, including a page by page counter or a main page only counter (total visits vs. page visits). Alternatively you can use an invisible counter, but I'd rather support a company with the button than hide the system all together.
After the data is collected there's a fantastic number of ways you can look at it thru the Stat Counter website. I've made the data of my blog entirely public, while you can't manage it, you can review all the stats of the page if your curious. Some fun features (fun more than useful) are a geographic locator of the people visiting your site powered by google maps. My strongest reader showings are in the US, Canada and the UK, with smaller showings in at least two dozen other countries. A more useful feature would be the keyword analysis which gives you a nice percentage based representation of the search terms used to reach your pages. There's also a most popular pages feature making it easy to figure out what you should write about more often (which I clearly ignore since I've yet to write anything else about QT4, despite it being my hands down most popular article).
To feed the curious and ego driven there's also a drill down tool that lets you look get more information about nearly anything on the site. This single semantic works across the whole site to provide more detailed information to the curious. It will go so far as to present a single visitor (with handy google map indicating where they connected from), what URL's they used to get to your site (for example a google search or cross site link).
The Blue Pill:
For those that need to feed their ego, track the stats of the site, or just curious where your users are coming from this plugin does the trick in spades. It's easy to setup, just drop the code into your page template and your up and running. It will also provide enough information for even the most discerning free service user.
The Red Pill:
One of the disadvantages of this system is the limited amount of overall data it allows you to store at 500 total entries. However this doesn't seem to play as much hard ball as you'd think, it allows 500 entries of full fidelity data, but all entries are considered for general counting and statistics. If you need more slots you can pay for them, but I've yet to figure out a reason why I would need them for a single page. For multiple pages I could see it being a great thing to invest in thou (as you can split your total alotment across all your projects however you please). Also this system is broken by people using things like google reader (as the stream has no java script capacity), and there are far more efficient ways to do this same functionality if you have access to the server the pages are being served from. That being said for a drop-in solution where you can't use a more efficient tracking method this works fantastic, and is generally accurate despite it's specific failings.