What we measure

We have created an analytics dashboard using Google Analytics and Google Search Console to gather information about how our users interact with the website and its content. This helps us to make improvements to the website to make sure we meet the needs of our users.

By using analytics data, we can measure:

  • pageviews (total and unique)
  • word count
  • average time on page
  • bounce rate
  • how users got to the page (for example, through a search engine, direct link or referral)
  • device type
  • most clicked sections in the table of contents or drop-down menu
  • Google search terms that brought users to the page

We have created an analytics dashboard that summarises these measures for business areas to use. For all of these measures, we can select the page and the time period we want to look at.

We also gather analytics data on downloads (PDF and datasets) and on-site search and navigation. You can request this by emailing content.design@ons.gov.uk

If you have any queries about the analytics data or measures used, or you would like to request additional insights into your content, email content.design@ons.gov.uk.

How to use the dashboard

The analytics dashboard has been designed to gather useful metrics about a web page with just a few simple pieces of information.

What you will need

To use the dashboard, you will need:

  • the full URL of the web page
  • the page path (the URL excluding “www.ons.gov.uk”)
  • the time period you want to look at

Creating a report

In the “Enter the page URL” field, paste the entire URL of the web page. For example, https://www.ons.gov.uk/economy/inflationandpriceindices/bulletins/housepriceindex/january2020. Set the drop-down menu to “Equals” to just include analytics data on the exact web page.

In the “Enter the page path” field, paste the URL excluding “https://www.ons.gov.uk”. For example, /economy/inflationandpriceindices/bulletins/housepriceindex/january2020. Set the drop-down menu to “Equals” to just include analytics data on the exact web page.

In the “Enter the date range” field, select the time period that you want the analytics to cover using the calendar tool. Once selected, use the apply button to select the time period.

The dashboard will update and refresh as you enter the information; you do not need to press a final submit to generate the report.

Saving and sharing

The report does not save when you leave or refresh the page, and you cannot link to the report for others to view. To share the report, download a PDF copy by selecting “Download report” using the arrow next to “Share” in the top-right.

Understanding the data

Pageviews

Pageviews lets us know how many users have looked at a particular page on the website. It provides total pageviews and unique pageviews.

Total pageviews includes users who visit the page more than once in the time period. Unique pageviews removes repeat visitors who may skew the figures and so provides a more accurate measure of page traffic.

Word count

We can measure the number of words included on a page and compare this with the amount of time users spent on the page. This gives us an idea of how much content a typical user would be able to read. 

For example, a typical user can read around 200 words per minute. If your content is 10,000 words long and the average user is only spending around three minutes on your page, they are only likely to read around 600 words, or less than 10% of your content.

Average time on page

The average time on page provides an average time in minutes and seconds that users spent looking at the page. The average time a user spends on a page on the Office for National Statistics (ONS) website is around four minutes. A typical user would be able to read around 900 words in that time.

Bounce rate

The bounce rate is the percentage of users who entered and left the website after viewing only that page. They did not interact with any other content or links on the site. Bounce rates should not be considered on their own, as a high or low bounce rate can be both positive and negative for different types of content and pages.

A high bounce rate is good for pages with a clear task, such as datasets, where the user comes to a page, completes the task (downloads the dataset) and leaves. If the bounce rate is low, it could suggest that the user landed on the wrong page or did not find what they were looking for.

A low bounce rate is good for pages that encourage further user journeys, such as product pages. A high bounce rate on these pages would indicate that users are not completing the journey to get to the content.

Device type

We can track the type of devices that are used to view the page. This includes: desktop, mobile and tablet.

Research tells us that mobile device users are more likely to be inquiring citizens, while expert analysts and technical users are more likely to use a desktop.

Our mobile users have been consistently increasing in recent years. On average, around 28% of website users are now on mobile devices.

How users get to the page

We can use analytics data to track where users come to the page from. This includes:

  • organic search: users who get to the page through search engines, such as Google and Yahoo!
  • direct: users who get to the page through a direct URL 
  • referral: users who get to the page by following a link on an external website, such as The Guardian
  • government sites: users who get to the page by following a link on other government sites, such as GOV.UK
  • social: users who get to the page by following a link on social media, such as Twitter or Facebook
  • GovDelivery: users who get to the page by following a link from a newsletter they received through GovDelivery (a web-based email subscription service)
  • email: users who get to the page by following a link from an email
  • other: any other way a user may get to the page or if the user’s cookies are turned off

Most clicked sections in the table of contents or drop-down menu

We can track what sections users are clicking on in the table of contents or selecting from the drop-down menu to help us understand the most popular sections on a page. These two actions are recorded using two separate tracking events; this is why some section headings appear in the most clicked section headings list twice.

Google search terms

Google Search Console provides us with a list of search terms that brought users to the web page.

Query

A query is a word or phrase entered into the Google search engine by a user.

URL clicks

URL clicks is the number of users who clicked through to the page from the Google search results for that query.

Impressions

Impressions is the number of users who saw a link to the page in the Google search results for that query.

If you would like to request any additional insights into your content, email content.design@ons.gov.uk.