Glossary of Dimensions and Metrics Used in Google Analytics
Legacy content
This page is for Universal Analytics (UA). Find a page like this for Google Analytics 4 (GA4)
EPA content related to Google Analytics is changing.
Google's legacy platform, Universal Analytics (UA), will reach end of life in mid-2023 with a one-time extension for contracting clients such as EPA until July 1, 2024. See KB article.
In these Web Analytics pages, content for Universal Analytics is marked "Google Universal Analytics (legacy)."
Content for the new platform, Google Analytics 4, is marked "Google Analytics 4 (GA4)."
- Dimensions are descriptive, like the name or URL of a page.
- Metrics are numerical, like pageviews or bounce rate.
- For example, in your Web Traffic Report, you see the "Primary Dimension" is Page (which displays URLs), and the Metrics are the numerical values that give you a picture of how each of those pages is being used:
GA has a very long list of Metrics, some of which are only useful for commercial sites. The glossary below defines the Metrics most commonly found in GA reports, or you can view the full list:
| Metric | Definition | Application |
|---|---|---|
| Bounces | A Bounce represents a Session with only one Pageview. | Whether a Bounce is negative or not depends entirely on the content of the Web page. Some Bounces occur because Users enter the site on the page that provides the content they want. However, you want to limit Bounces from pages that are meant only to direct Users to other pages on the website. |
| Bounce Rate | The Bounce Rate represents the percentage of single-page Sessions (Total Bounces divided by total Sessions). | See Bounces |
| Entrances | Entrances represent the number of Sessions that started on a specific Web page or group of Web pages. | Entrances can show you how often specific Web pages serve as Landing Pages for Users. NOTE: if you mistakenly apply Sessions to the Page Dimension, the data will actually represent Entrances, rather than Sessions, since Sessions cannot be applied at the page-level. |
| Events | Events represent custom actions that are distinct to a particular GA account. Events can include actions like clicking hyperlinks, images, or stop and play buttons in video players. | EPA tracks clicks on
|
| Exits | Exits represent the number of Sessions that ended on a specific Web page or group of Web pages. | Exits are distinct from Bounces: Exits occur in all Sessions, but Bounces denote only single page Sessions. You want to limit Exits on pages that are meant only to direct Users to other pages on the website. |
| New Sessions | New Sessions represent the number of first-time Sessions during a specified time period. The presence of cookies in the Web browser determines whether it is a New or Return Session. | New and Return Sessions can be used as Metrics or Dimensions (i.e. User Type) to help you understand how User behavior differs across your website. Remember that some Users delete cookies and use multiple devices. Both of these factors can inflate New Session counts. |
| Pageviews | A Pageview represents an HTML page being loaded in a Web browser. Multiple views of the same Web page are counted in this Metric. Non-HTML files, such as PDF documents, are typically counted using Events, rather than Pageviews. | Pageviews measure traffic volume over time at the page-level, whereas Sessions measure traffic at the Session, or Visit, level. You can use Pageviews to measure aggregate traffic for groups of Web pages and subsites. |
| Page Value | Page Value is used in e-commerce to assign a relative value to any page viewed prior to a monetary transaction. | Do not use them. |
| Pages per Session | Pages Per Session represents the average number of Pageviews during the course of a Session (Total Pageviews divided by Total Sessions). | Pages per Session measures the depth of a Session and can be a measure of engagement, depending on the content of your website. Pages per Session is an average, and should be measured across different segments to help understand User behavior. |
| Percentage Search Refinements | % Search Refinements represent the percentage of internal search queries that resulted in at least one additional search in the same Session. | % Search Refinements can be an indicator of Users not finding what they need in one search attempt. Use % Search Refinements carefully, though, because some Users may perform multiple searches for different information during the same Session. |
| Return Sessions | Return Sessions represent the number of repeat Sessions during a specified time period. The presence of cookies in the Web browser determines whether it is a New or Return Session. | See New Sessions |
| Sessions | A Session represents the full timespan that a User spends on a particular website, starting when the first Web page of that site is loaded in the browser and ending when the User either leaves the website, closes their browser, or the Session times out after 30 minutes of inactivity. | Sessions measure website traffic volume over time and form the foundation for many important Web traffic metrics. Since Sessions encompass the full range of activity Users perform while on a website, they are not typically used to measure single Web pages or subsites. |
| Session Duration | Session Duration represents the average amount of time, in seconds, of a Session. Technically, it represents the time between the start time of the first Pageview and the start time of the last Pageview or Event. | If the final Pageview of a Session does not include an Event, like an external link click, there is no way to calculate the amount of time spent on that page. Similarly, Sessions with only one Pageview and no Events (see Bounces) will have a Session Duration of zero seconds. Consider measuring Session Duration for Sessions of two or more Pageviews. |
| Sessions with Search | Sessions with Search represent the number of Sessions in which the User utilized EPA's internal search tool. A successful search is recorded each time the Search Results page loads and the URL string includes the parameter "querytext." | Sessions with Search is an important Metric because it shows the percentage of Users that access EPA search as part of their Session. Compare this Metric across different segments to help understand what audiences are most likely to use internal search. |
| Time on Page | Time on Page represents the average amount of time, in seconds, a User spends on a particular page. Technically, it represents the time between the start time of a given Pageview and the start time of the subsequent Pageview or Event. | EPA tracks external links as Events, so they are used in Time on Page calculations. However, Sessions that include only one Pageview and no subsequent Events (see Bounces), will have a Time on Page of zero seconds. Zeros will lower the average, so Time on Page should be used carefully. |
| Unique Events | Unique Events represent the number of Sessions during which the specified Event (e.g. PDF download), or group of Events, was triggered at least once. Duplicated Events are not included in this Metric. | Users can trigger the same Event countless times in a single Session, but it will count as only one Unique Event. Compare Total Events to Unique Events to analyze how Users interact with different Web pages. |
| Unique Pageviews | Unique Pageviews represent the number of Sessions during which the specified page, or group of pages, was viewed at least once. Multiple views of the same page and page refreshes are not included in this Metric. | Unique Pageviews is an important metric because it functions similar to Sessions, only at the page-level. Each Web page can garner many Pageviews per Session, but only one Unique Pageview per Session. Compare Total Pageviews to Unique Pageviews to analyze how Users interact with different Web pages. |
| Unique Users | Unique Users represent the number of unique Web browsers that access a website during a specified time period. The presence of cookies in those browsers determines whether they have accessed the site before. Ideally, each browser is counted as a Unique User only once in a given time period. | Unique Users should not be applied with short time-periods since most Users will be unique over the course of a few days. Remember that some Users delete cookies and use multiple devices and browsers. These factors can inflate Unique User counts. Like most Metrics, it is most helpful for long-term trend analysis. |