This step involves defining what pages and journeys will be tested.
By using a structured approach, you can help ensure that all parts of your site are compliant.
The size of the sample will be dictated by the size and complexity of your site. A single-page application may have the entire site/journey tested, but a large site with thousands of pages will have a sample tested.
In the structured sample, you should include pages which include representative content identified in Step 2 exploration.
This means that there should be at least one page which contains every type of:
You should also include each of the critical journeys identified, or a sample of these if there are too many.
This means you will test the end-to-end journey that a user is expected to complete (such as submitting a return, or applying for a service).
Each test in this instance may be scope limited in order to align with your internal areas of responsibility, but this will be made clear, as will external failure dependencies (such as the site in scope test complies, but users must use a non-compliant site in order to complete their task, making it an end-to-end failure)
Where the structured sample doesn't cover the entire site, you should add a randomised sample of pages to be tested.
You should test random pages equal to 10% of the number of structured pages checked, or at least 10, whichever is higher.
On some very small sites, or sites where the structured tests represent a very high proportion, it may not be appropriate to add a random sample.