Metrics We Track



  • Page views

  • Uniques

  • Average/total engaged time

  • Average scroll depth

  • Social actions

  • Device type

  • Referral buckets


  • Page View Score

Cool, but what does all of this actually mean? Allow us to explain...



Page views are a count of the number of times our javascript code was loaded on a certain content item. We do not de-duplicate page views on a time frame or user basis. If a user loads a content item and reloads the page 10 seconds later, we count that as 2 separate page views.


Uniques provide the number of individuals that access the content, and are de-duplicated based on the dashboard. We count uniques via a cookie that does not include any personally identifiable information. We currently don't de-duplicate uniques across browsers or devices. In early 2018 we updated our method for calculating Uniques to count users in a more efficient way, while maintaining high accuracy and error rates below 3%.


Average and Total Engaged Time are calculated as active time with content. Active time means the content must be in the current active tab in the browser, the browser must be the focused window, and the user is performing actions (scrolling, moving their mouse, etc.). We continue to count time for the duration of the page session, which we cap at 20 minutes. If a user leaves their computer for 20 minutes and re-engages with the content, we will start counting again, but will not count the 20 minutes.

If you're looking at those numbers in a export, average engaged time x page views = total engaged time in seconds. 


Average scroll depth gives the percentage of page views that reached a certain point in the content. We capture the data by quartiles (25%, 50%, 75%, 100%). Each subsequent quartile is a subset of the previous one, meaning, if someone hits 25%, and then 50%, they will be counted in both.


Social actions are an endorsement or share of content via a social network or aggregator. We work with the social networks’ APIs to capture all social action data based on the URL of the content. Our system knows to start looking for social actions as soon as a piece of content receives one page view with our tag active on the page. Social actions include:

  • Facebook reactions, comments, and shares

  • Pinterest pins


Referral buckets are determined by utilizing the referring URL in the header of the HTML. This is a standard method of determining referral source. We break out referral source into 5 major buckets: social, apps/IMs/email, search, internal, and other sites.

Social: reflective of any social networks or aggregators we track. For example, a user on Facebook who clicks on a link to your content and then arrives on your content, is counted as one social referral.
Apps/IMs/email/etc.: traditionally called “Direct” traffic, signifies no referring URL.Also contains traffic from web-based email tools like Gmail or Outlook Web.
Search: major search engines, in aggregate (i.e. Google, Bing, etc.).
Internal: internal traffic from your site.
Other sites: known sites that do not fall into any other buckets.

We can also determine whether referrals are organic or paid using a combination of query string parameters and referring URL.

If a piece of content is shared with the query string parameters still included in the link, we will continue to count any traffic driven from that link as paid traffic. Effectively, paid traffic accounts for both paid and earned visits to content.


Device type is determined by utilizing the user-agent ID of the browser that is being used to read the content piece. Because browsers are unique to devices (i.e. Chrome for desktop vs. Chrome for mobile), we are able to determine which device type is being used. We break out device type into desktop, mobile, and tablet.


Page View SCORE

The Page View Score uses a 0-99 index based on expected traffic over the next hour to create a snapshot of content performance right now. This is ideal for stack ranking your content items and understanding their relative performance against one another.