4 GOOGLE ANALYTICS TRACKING MISTAKES (AND HOW TO FIX THEM)

<p>#analytics #tracking #google</p> <p>There’s a high risk of making bad decisions if you’re blindly relying on the data that you see in #google #analytics.</p> <p>Hi, i’m cally, from the international institute of digital marketing. Here are 4 tips to help you minimize data skewing factors.</p> <h2><br></h2> <h2><strong>1. Missing or duplicate #google #analytics code</strong></h2> <p>This sounds trivial, but it’s a common problem—especially on sites that use more than one cms.</p> <p>The good news is that #google #analytics has missing code notifications built-in. The bad news is that it’s slow, and may take weeks to alert you about pages with missing code. It also doesn’t tell you about duplicate codes, which is another common problem.</p> <p>For that reason, it’s best not to rely on #google’s notifications and instead crawl your site for errors with a tool that allows custom extraction.</p> <h2><br></h2> <h2><strong>2. Incorrectly set up interaction events</strong></h2> <p>If you’re using interaction events for #tracking events that fire automatically on each page, like scroll depth #tracking, that’ll result in close-to-zero bounce rates across your whole website—which isn’t good.</p> <p>You can easily spot these issues by looking for unrealistically low bounce rates in ga.</p> <p>If you suspect interaction events as the culprit, change the event’s “non-interaction hit” setting from false to true in #google tag manager.</p> <h2><br></h2> <h2><strong>3. #tracking hits from other domains</strong></h2> <p>It’s surprisingly easy for someone to mess up your data if you don’t take preventive measures. Why? Because your ga/gtm #tracking code is visible to anyone who opens your source code, so anyone can send hits to ga servers under your #tracking code.</p> <p>Luckily, it’s easy to prevent this from happening by setting up a view filter.</p> <h2><strong>4. #tracking your own sessions</strong></h2> <p>There are a lot of irregular actions we make on our websites, and we don’t want those reflected in our data.</p> <p>So