You can measure the performance of your content by using metrics such as average attention time, engagement rate, and scroll depth.
Measure engaged content
When you read a piece of content, you are considered engaged if your page is in focus and you scrolled or moved your cursor in the past 5 seconds. After 5 seconds of not moving your mouse, or not looking at the page (perhaps it is in another window or tab), Optimizely Content Marketing Platform (CMP) takes the current time and subtracts it from the time you started being engaged.
For example, if you loaded the page, scrolled or moved your mouse for 7 seconds, then stopped, your total engaged time is 12 seconds because you were active for 7, and 5 additional seconds is assumed.
Why is there a 5-second timeout? Chartbeat found that if someone is inactive for more than 5 seconds then that means they weren't looking at the screen.
Therefore, if you come back and start reading, your cumulative engaged time is added to the first 12 seconds. If you decide to then close the window, your total cumulative engaged time is recorded.
Measure average attention time
Average attention time is calculated by adding the engaged time of all users and dividing the sum by the number of pageviews. This gives an average of how much time was spent on a particular page. The higher the average attention time the better, as this indicates that people spent more time reading your content.
Measure scroll depth
The scroll depth tracker measures the amount of the page a reader has seen, with the top of the page defined as 0% and the bottom of the article (not including footer) as 100%. As a user scrolls down the page, multiple tracking events are fired at each 10% interval to capture the maximum depth viewed. These events automatically account for screen width and magnification.
For example, in the diagram below, the tracker would register a 50% scroll depth.
You can view these metrics by article.
You can also click into an article and scroll to the bottom to see a chart and compare it to the average scroll depth of articles with similar length. This is calculated based on all-time data of articles within 100 words of similar length.