A strategy is a hypothesis: based on a strategic diagnosis of the situation, it proposes actions that are believed to create an advantage for the organization.
Because strategy is (generally) produced under conditions of uncertainty, it cannot be generated through deductive reasoning alone. Rather, it requires abductive reasoning to produce the key insights that emerge through effective sensemaking and diagnosis. Deduction cannot produce new ideas; it assumes a computational model of reality that does not line up with the kind of radical uncertainty that characterizes the world.
This is not to say that strategy is not empirical; on the contrary, any good strategy is but a hypothesis that can be tested. But arriving at that hypothesis takes more than deduction.