Web analytics are tricky.
Not trick-or-treat tricky, but deceptive-like-a-con-artist tricky.
Web analytics can tell you one thing and mean another. Or mean nothing at all.
Traditional devices like straw polls and cold-call surveys have been replaced by nearly real-time (or in some cases, totally real-time) statistics via applications like Google Analytics.
This simple blog gives me information such as which posts have been the most popular, how many page views I’ve had each day and which blogs have linked in to mine.
We use analytics in the Missourian to figure out what people are reading – and what they aren’t. The numbers inform our homepage decisions and spawn article ideas.
But as NYT executive editor Bill Keller points out, “We don’t let metrics dictate our assignments…because we believe readers come to us for our judgment, not the judgment of the crowd. We’re not ‘American Idol.’”
Online data are useful. Running track and field for ten years has ingrained in me a deep appreciation for numbers, stats and performance metrics.
Involvement in a sport that thrives on data points down to the centimeters and hundredths of seconds also instills a knowledge that metrics are misleading.
Examining a spreadsheet of race times does not reveal a bout of the flu or a wicked headwind. Sifting through web metrics does not always expose how many devices one person used, for example. Three unique hits can originate from one person: personal computer, work computer, iPhone.
The take-away is to handle web analytics with a few shakes of salt and a critical eye. One measure I particularly enjoy is how long readers stay with a story. Again, this stat could be inflated by an act as simple as leaving a web page open and walking to the kitchen to make dinner.
In my Vox article about the Cycle Chic movement, readers stayed on the page for nearly two minutes. Translation: most read the entire article.
That’s one metric I can feel good about.