Pogo-sticking dashboard
Pogo-sticking is when a user quickly navigates back and forth from a web page within the same user session - for example, when a user navigates from Page A to Page B back to Page A and then to Page C.
There are 2 types of pogo-sticking:
- internal, which is navigating quickly back and forth from a GOV.UK page to other GOV.UK pages
- external, which is navigating quickly back and forth from an external search result page, such as Google, to GOV.UK pages
The pogo-sticking dashboard shows the internal and external pogo-sticking rate for GOV.UK pages. This dashboard does not track when users pogo-stick between different smart answer results.
Using the Pogo-sticking dashboard
Get access to the Pogo-sticking dashboard
You can view the pogo-sticking dashboard if you have a Google account which uses the @digital.cabinet-office.gov.uk email domain.
Selecting the date range and page path
The dashboard is split into internal and external pogo-sticking.
These sections have the same metrics for both internal and external pogo-sticking.
To get pogo-sticking metrics, you must specify 2 variables.
Select a date or date range in the top right hand corner.
Enter a page path in the Page equals field. The input text must match the page path exactly, and URLs must start with the domain, e.g.
www.gov.uk/find-a-job.
Within the internal and external sections, these sections are split between the original page of interest, and a page to compare that original page to.
You can also enter a page path in the Page equals field under Comparisons with another page.
You can now view the pogo-sticking metrics.
If you have any issues with the dashboard, email the Analytics team.
Metrics
Total number of sessions
The total number of sessions for the particular date or date range for desktop and mobile devices.
This metric is in both the original page and comparison page sections.
Percentage of the sessions where the page triggered pogo-sticking
The percentage of sessions that visited a page where that page triggered pogo-sticking for Desktop and Mobile devices for a particular date or date range, averaged over the dates.
For example, if the total number of sessions that visited a page is 1000, and the % of those sessions where that page triggered pogo-sticking is 45%, then pogo-sticking occurred during 450 out of the relevant 1000 sessions.
This metric is in both the original page and comparison page sections.
Time-series graph
This graph plots the percentage rate of the sessions that visited a page where that page triggered pogo-sticking for Desktop and Mobile over the past 28 days. Hover your cursor over the graph to see the percentages.
Common pogo-sticking sequences
This table displays the most common pogo-sticking sequences over a date or date range.
Count is in descending order. Desktop and mobile journeys are added together.
Comparisons with other document types that triggered pogo-sticking
There are 3 plotted graphs in the Comparisons with other document types that triggered pogo-sticking section.
Each plot represents a different document type: guide, transaction and step-by-step navigation. These plots include all page paths of the specified document type.
Under the document type, a table presents the following pogo-sticking rate metrics:
- minimum
- 25th percentile
- median
- 75th percentile
- maximum
Using the Guide plot as an example, if the:
- minimum pogo-sticking rate is 2%, this means that the lowest pogo-sticking rate of any page path characterised as a Guide is 2%
- 25th percentile is 16%, this means that 25% of page paths characterised as Guide have a pogo-sticking rate less than 16%
The plot displays the quartile ranges (lower-, inter-, upper-).
The black dot and value represent the average pogo-sticking rate for the page path of interest.
Desktop and mobile pogo-sticking rates are averaged together.
How the Pogo-sticking dashboard works
Data sources
The Pogo-sticking dashboard uses custom queries to access gds-bq-data.GA4_PogoSticking tables stored in BigQuery.
These tables are constructred using a number of scheduled queries.