"Decisions reorganize the future more than they solve the present.
Most disasters begin as a reasonable exception. In a migration that looks reversible until the data moves, the system learns the exception faster than it learns the risk.
In a migration that looks reversible until the data moves, you will feel pressure to keep the conversation optimistic. Optimism is expensive when it prevents guardrails. You see it in a local fix that creates a new problem at the seams. Define the decision rights before the crisis, or the crisis will define them for you. If the exception path is informal, it becomes politics; write the signal that makes an exception legitimate and who can approve it.
James G. March is a reminder to audit the argument, not the intention. Write the failure story forward while the decision is still reversible, then insist on one guardrail. Stop while it is still simple.