How do I bring a heavily customised ITSM platform back to baseline?

Heavy ITSM customisation usually starts as helpful exceptions and ends as upgrade risk. The practical path back to baseline is to freeze new custom work, inventory every deviation against a clean reference instance, and keep only what a named owner can defend with a real business outcome. Fix process rules, state models and fulfilment flows before you tidy the workspace layer, because the workspace only presents what the platform already does. Prefer deactivating unused pages and rules over clever rewrites, and keep a short exceptions list with expiry dates. If most deviations fail the outcome test, a controlled greenfield rebuild can cost less than years of overlays. Success is a rehearsed upgrade and a week of live work without people falling back to old screens. Challenge habit as hard as technology. Ask who would notice if each custom field or state vanished tomorrow. If the answer is thin, remove it. John Sansbury guidance in The Journey to Service Excellence favours simpler out of the box designs for cost, upgradeability and reliability, and warns that complexity makes systems harder to maintain and repair.