Analytics & Testing, Content Marketing

Google’s Bogus URL Shortener Statistics

We had an intriguing session with a client as part of some analytics training and consultation we’re doing with their parent company. As part of their ongoing efforts, they distribute QR Codes, append Google Analytics campaign code, then apply the Google URL shortener, allowing them to accurately measure the response rates of their efforts.

This is a solid strategy. Analytics alone can’t provide you with everything you need because of all the applications that distribute links nowadays that don’t append referring data to the request. Analytics knows where you came from by web servers telling the next page where the last page was along with all the information on the client. Since apps don’t have a web server… they don’t pass data. As a result, you’ll want to append campaign code to your shortened URLs before distributing them. We just showed how to add Google Analytics campaign code tracking withHootsuite recently.

All was well with the world until the Marketing Coordinator decided to dig into the Google URL shortener stats a bit more. He simply couldn’t get the numbers to come close with what Google Analytics was providing. There are several reasons for this… Google URL Shortener is a pass through service that measures every single click, while Google Analytics is a JavaScript based solution that provides the majority of the data.

However, he discovered something much more alarming… he found that Google had clicks occurring before the link was ever created! Here’s proof – directly in Goo.gl’s reporting engine:

And it didn’t happen just once… it’s all over the place!

It’s unfortunate that this data can’t be relied upon… but it simply can’t be. Through the interface, you can’t select date ranges, so Kevin has to manually drag his mouse across the charts to capture dates and clicks to fill out his reporting. I’m surprised that Google doesn’t simply incorporate their shortener in with their Analytics and automatically register the campaigns. I’m more surprised, in this day and age where we need to research our site performance, at what a blunder this error is!

Anyone know a product manager over at Goo.gl that can explain this?

One comment

  1. 1

Leave a Reply