Patent Law

Using Google AdWords For Site Statistics

By Erik J. Heels

Name the one stat that matters.

I have lots of sources for website stats:

1. Apache log files, periodically processed (via cron jobs) with Urchin
2. Lijit search stats
3. Google Analytics
4. FeedBurner feed and site stats
5. Google AdSense revenue

Raw log files are fun, but with so much junk traffic out there (click fraud farms, referrer log spam generators, contact form spam generators), the majority of my raw website traffic is fraudulent. Or at least non-human. And that's not even counting the legitimate crawlers, bots, and the like that are indexing my site for search engines.

The stat sources I use most frequently are Google Analytics and FeedBurner (also owned by Google).

I found that I was checking my server log files rarely. Plus they are a pain to maintain (rotating, archiving, cron jobs, and the like). So I turned off logging on all of my websites and deleted all of the log files. It was cathartic. Kind of like shredding my paper files from the 1990s.

At the same time, I turned on Google AdWords both on my site and in my feed. Not because I care about the money, but because the amount of money generated is a good measure of the overall effectiveness of my content (notice I didn't say "website").

Creating content that generates a ton traffic is all well and good. But if you can't monetize (oh how I hate that word) that traffic, what's the point? A lot of business from the 1990s failed because, quite simply, they failed to sell anything. You can't eat ego.

Which is also why selling women's shoes (and blogging about shoes) is so satisfying. The write-click-sell cycle works a lot better with tangible products than with intangible services.

So how much time are you spending on stats? And why?

Full post as published by Erik J. Heels on February 25, 2008 (boomark / email).