Knowledge transfer

Turn documentation into an independent handover.

Use practical exercises to check that another person can deploy, diagnose and recover the system.

Explore the playbookDownload PDF
On this page

A handover rehearsal

  1. 1

    Deploy a change

    The receiving maintainer follows the documented path

  2. 2

    Diagnose a failure

    They find the relevant trace and identify the affected dependency

  3. 3

    Restore a workflow

    They recover the system and verify a business operation

Prepare a safe environment

Use representative data and controlled failures. Make access, credentials and dependencies available before the session begins.

Let the receiving team drive

Observe where the instructions or system are unclear. Update the runbook and repeat the task until the gap is resolved.

Agree what ownership includes

A handover should reflect the responsibilities the receiving team will actually take on. Maintaining an application requires repository access, a reproducible build, an understanding of environment configuration and a way to confirm that a release is healthy. Agree which responsibilities transfer and which remain with another supplier before scheduling the sessions.

List the systems, accounts and documents involved, with an owner for each. Check that instructions match the receiving team's tools and permissions. A deployment guide that assumes an account only the original developer possesses is an unresolved dependency. Include commercial dependencies such as subscription ownership and renewal contacts, without placing passwords or recovery codes in the handover document.

Rehearse a change and a failure

Choose a small change that follows the normal delivery path. Ask a receiving maintainer to create it, run the checks, deploy to a non-production environment and confirm the result. The original team can observe and answer questions, but should not silently take over. Record every missing permission, undocumented command and assumption that interrupts the exercise.

Then introduce a bounded failure in a test environment. Ask the maintainer to identify the affected workflow, find the relevant trace and explain the safe next action. They should be able to distinguish an application fault from an unavailable external dependency. A realistic, modest exercise often reveals more useful gaps than a presentation about a hypothetical major outage.

Close the handover with evidence

Turn the rehearsal findings into assigned actions. Update the runbook, correct access and repeat any task that could not be completed independently. Agree where the receiving team should ask for help and the availability of the outgoing team. A handover does not create an unlimited support obligation unless that has been agreed separately.

Keep a record of the application version, environments covered, participants and tasks completed. List known limitations and the owner of each unresolved issue. Make operational documents easy to find from the repository or service catalogue. Successful knowledge transfer means the receiving team can carry out its agreed responsibilities without depending on undocumented knowledge held by one person.

An updated system map, task-based runbooks, access ownership and a record of unresolved questions with named owners.