Source: https://moz.com/blog/misuses-4-google-analytics-metrics-debunked
These are some commonly used metrics in Google analytics that are often misinterpreted. The key is to understand how the metrics work to be able to avoid making assumption based on collected data.
Misunderstood metrics in Google Analytics
1. Average time on page – this data can be misleading since Google uses the next page to compute the time a user spent on the current page. This means that a user entered your site and spends time reading then left your site is equal to zero session duration.
2. Site speed – this report gives you visitor information (location, browser, OS, etc.) and how users interact with your page. This report also includes page-load time, image load speed, button click response and information that can help improve your website’s user experience. Keep in mind that outliers can create a very large influence on your data. You can increase sampling rate to see performance of new pages that has less traffic.
3. Conversion rate – this can often mislead your data a visitor conversion can fall under organic traffic data if they: click your site from Google.com (organic), then revisit your site thru “direct” traffic then successfully finished a conversion funnel. The solution to this is by using last interaction model in attributions.
4. Exit rate – the thing about exit rate is that sometimes webmasters would automatically assume that a landing page that garnered a high exit rate in Google Analytics means a problem to the website. Some pages normally receive high exit rates because they are located at the final pages of a checkout funnel.