Agree what a software update must achieve before building it
A software update is described as improving the dashboard, but nobody has specified what a user should be able to do differently or how errors will be…
The Kay & Co. resource library
Website changes & recovery
Check a website update before it goes live and agree how to put things right if a change causes problems.
Find an answer
Search a question, a task or a few words.
Search results
Matching resources
A useful starting point
Start here
Publish the version your practice actually approved
The practice approves one preview, but the supplier publishes a later build containing additional unreviewed changes.
Include external settings when planning a rollback
The previous website version returns, but an external configuration still points to the new incompatible behaviour.
The complete list
All website changes & recovery resources
Showing 1–24 of 53 resources, ordered by title.
Agree how much downtime and lost work your recovery plan allows
A backup runs nightly, but the organisation has never agreed how much recent work it could lose or how long a service could be unavailable. Backup frequency…
Agree what a software update must achieve before building it
A software update is described as improving the dashboard, but nobody has specified what a user should be able to do differently or how errors will be…
Agree who is responsible for your website launch
A launch often involves more than one supplier and several people at the practice. Name the person who decides whether the site is ready and make sure they…
Carry emergency fixes back into normal publishing
A supplier edits the live output directly, and the next routine release silently restores the original fault.
Check deeper guides and downloads after a restore
The recovery team declares success when the root page loads, while important articles and files remain absent.
Check every important page type affected by a shared change
A shared dependency changes, but the sample checked for unintended effects excludes other pages using it.
Check live services without putting secrets in release reports
An environment setting is copied into the release and public functionality points to a test service. Validate environment-specific destinations without…
Check that a healthy link report describes the live website
An automated audit reports healthy links because it resolves relative routes against a preview website domain. The production release may contain different…
Check the live service before closing a software release
An update is deployed and the ticket closes immediately. No one verifies the live enquiry journey or records which version is running. A successful…
Check the public delivery of a rollback, not only its source
The original deployment is restored, yet some visitors continue receiving the faulty response from a cache layer.
Check your backups cover the working service
A team backs up website files but has no recoverable copy of form configuration or the records needed to rebuild its publishing process. A backup needs to…
Check your team can communicate during an outage
The recovery plan tells staff to use a messaging platform that is itself unavailable in the chosen scenario. An outage can affect the messaging tool named…
Choose a realistic and bounded recovery exercise
A recovery exercise assumes every administrator, system and instruction is available. It misses the obstacles that could make a real outage difficult.…
Choose a release time with the right support available
A planned website change overlaps with a major campaign launch and the only available support engineer's leave. The technical task is small but recovery…
Close or protect temporary recovery websites after the drill
The exercise succeeds, but the temporary restored copy stays accessible and begins circulating as another version of the site.
Combine suppliers' changes before either overwrites the website
Different suppliers publish independent updates, and the later deployment removes changes accepted in the earlier one.
Confirm everyday work is possible after recovery
A system is technically restored, but users cannot complete the essential workflow because permissions and a dependent integration were not checked. A…
Confirm the release reached the public domain
The hosting dashboard reports success, while the organisation's main domain still serves the previous release.
Decide what to keep before redesigning your website
A redesign is an opportunity to improve useful content, but it is easy to lose pages that staff or patients rely on. Begin with a list of current pages and…
Fix a section that keeps sending visitors round in circles
The website opens normally except for a section that continually redirects. Two routing layers may contradict each other.
Give a website move one owner across suppliers
One supplier changes the domain while another controls redirects. Neither verifies the complete incoming journey.
Give recovery a tested backup administrator
The incident owner knows which deployment to restore, but the only authorised account holder cannot be reached.
Give temporary campaign redirects an end date or trigger
A short-lived routing change remains after its campaign ends and masks the original destination. A temporary campaign rule needs an exit condition, not just…
Include external settings when planning a rollback
The previous website version returns, but an external configuration still points to the new incompatible behaviour.