Here’s why I decided to create my own web traffic log analysis tool:
That’s the graph that my site hosting company is providing me, showing traffic to my site in the form of daily “page views.” Those spikes are traffic from me, making changes and updates to the site, and then testing the public view. Those page views are generated by JavaScript, so it doesn’t show me traffic from users without scripting enabled. More importantly, this data doesn’t help me understand just how many people are visiting certain areas and resources on the Globalisk site.
The “detailed statistics” link doesn’t help much, either. It’s also page views, but by directory/page, and without breaking the total down by day/date. Some numbers look terribly wrong to me, too.
Although I did manage to find a perfectly good, free log analysis tool for Windows, it produces too many reports, detailing too much information I don’t want, and no view showing exactly what I did want. Since I’ve done plenty of NCSA log processing in the past, for a couple different websites, the simplest solution seemed to be my favorite part about being a programmer: make my own app.
NCSAnalysis is a simple, free, Java-based web analysis tool that processes log files to produce Excel-formatted reports detailing unique IP (site visitor) traffic on a daily basis:
Excel view of data produced by NCSAnalysis
You can filter out traffic from certain IPs, and from web bots/crawlers. It also lets you track hits to individual resources, such as program installation files, audio files, etc.
OBD
