Can You Trust the Accuracy of Google Analytics?

 

There's been a lot of talk around the blogosphere recently about the accuracy of Google Analytics. I'm personally not a fan and I use my server stats as ultimately it is a lot more accurate.

There's no question that GA is one of the world's most widely used statistics services. It is free, easy to install and fairly simple to use but it would be a mistake to assume it is perfectly accurate. Most bloggers that I know of use GA as do a lot of businesses and online magazines. It offers a wealth of information on your visitors from traffic sources to unique visitor counts.

The main problem with GA is that it uses javascript to track visitors. Now most internet users use javascript but it is estimated that 10% of users have JS deactivated. So on a site with 50,000 visitors a month it equals a discrepancy of around 5000.

The tracking of unique visitors is also inaccurate as some users may visit from several different devices. They could view your site on their laptop at home, or their iPhone or even a work computer. So the unique visitors statistic may be higher than the actual amount of unique visitors. This discrepancy could be large or small depending on the amount of visitors who use multiple devices.

Like Alexa, Google Analytics uses sampling and while this method is widely used and mostly accurate it can produce discrepancies.

My personal experience is that Google Analytics regularly under reports my stats. For example one month my server stats showed I received 50000 pageviews on my fashion blog but GA showed stats of 9000 pageviews for the same month and that's a drop of more than 80%. Speaking with other bloggers I've realised that a massive drop is fairly common.

So in answer to the question posed in my post title I'd say NO! Google Analytics is about as far from accurate as you can get.