After the community of agents solve a problem, each individual agent's execution trace is prepared for storage into collective memory. Performance improvements can be gained that are not possible without preparation. The execution trace is prepared in order to remove inefficiencies and unimportant details and to segment it into ideas. A sequence of actions may be beneficially resequenced; preparation exploits this property of the knowledge representation. We will measure the benefits by showing that preparation improves upon adaptation and/or reuse alone.