Why analytics matters.
“Half the money I spent on advertising is wasted. The trouble is, I don’t know which half,” said John Wanamaker, over a century ago.
Today, web analytics allows firms to track ad dollars with a precision never before possible. But absent solid statistical analysis, raw numbers can be misleading.
The figure above is an illustration of why analytics matters. Two web ads are tracked for their impressions and click-through rate. While ad B has almost twice the click-through rate of ad A (4.3% versus 2.4%), we have different amounts of information about both. In short, using a Bayesian approach, we can say we believe there’s just a 66% chance that ad B performs better. A difference, but not necessary significant enough to act on.
Data visualization and statistical analysis allow us to go beyond simple averages, and understand the complexity of our data.
