Five days to launch

Engineering estimates three days for a manual workflow and three weeks for automation. The first pilot has twelve accounts.

Consider the pilot's purpose, operational capacity, errors, demand uncertainty, and the cost of delay.

A manual path may fit a small learning pilot if a named team can perform the work reliably. Quantify expected effort, add checks, and set a volume ceiling with a review trigger.

If the action is high risk or volume is already large, the shortcut may be unsuitable. A deadline does not remove operating constraints; it forces an explicit choice about scope, capacity, or timing.

Make the hidden manual work and the transition condition part of the recommendation.