I have provided an approach to investigating how planning should take place for a unitary organization. In particular, it has been proved that given an information structure of multiple, independent random variables, a best action in combination of goal setting should maximize the conditional expected payoff, and that the planner should search for information structures that are at least as valuable as the information structure available. The case of planning with goal setting in terms of selecting a probability density function for the states in the grand world is interpreted in a similar way. No consideration is given in the current formulation to the cost of information. The logic of planning developed here is based on the criterion of maximizing expected utility. Such normative interpretation of human behavior has been criticized for being unrealistic (e.g. Dawes, 1988). The concepts developed here are thus subject to the same criticism. The proposed explanation of planning behavior can, however, enhance our understanding of how planning should and does take place for a unitary organization. The current formulation can be extended into multiple organizations with conflicting interests in the future.