The GERALD Approach
Four stages. No acronym.
The GERALD Method is a proprietary presentation of ordinary, useful decision discipline. Each stage saves independently, supports revisions and identifies missing information. Solutions is a division. Solving is a later, separate finding, and usually not the finding.
01
Understanding the Environment
Start with the actual decision question, the linked CARL problem when one exists, frozen criteria, constraints and available evidence. Missing information stays missing.
02
Establishing Direction
Compare user-defined options against explicit weights. Scores are entered by people or marked as estimates. A weighted result is a decision aid, not objective proof.
03
Enabling Delivery
A human selects the action and records rationale. Choosing an option creates a draft proposal, not an approved implementation. AI cannot approve its own suggestion.
04
Recognizing Outcomes
Implementation is an event. Effectiveness is a separate, evidence-backed finding against each linked criterion. Partial results stay partial. A named solution may finish here still unsolved.
A live solution cannot enter implementation until it is connected to a same-workspace problem. Templates may wait. Implementations may not wait for a definition of success. Completing the named work still does not close the CARL matter.
Open the Method workspace