A good strategy is not always successful, but even an "inappropriate" strategy may be an actual strategy. A "bad strategy" is one that doesn't even try to address an important challenge. Instead, it speaks of aspirations, visions of the future, lays out performance goals, or simply lists a bunch of unconnected actions.

A real strategy is a coherent mix of policy and action designed to overcome a significant challenge. So a sensible employee might indeed say that they have no idea what the organization's strategy is - because it seems to have none. Senior managers' so-called "strategies" are heavy with aspirations and goals, but light on how resources and strengths will be combined to achieve them.

The US needs military/defense, economic, and social strategies. A medium-sized business, by contrast, is normally best off focusing its efforts on a single crucial objective.

A real strategy is not bottom up because it deals with issues that require unexpected or unusual types action, especially of coordination among units.

A good strategy focuses efforts on a target, and that focus can only be achieved by not diffusing energy in other directions - that is the meaning of Michael Porter's dictum of "choosing what not to do." At the same time, a good strategy chooses the right target to focus on, not wasting the focus of energy on a target that cannot be affected or that is unimportant - that is the meaning of Drucker's distinction between efficiency and effectiveness.

I agree with O'Toole that custom and comfort are impediments to change. However, it is important to recognize that resistance to change is logical as well. The new "change masters" literature seems to take change as the norm. It isn't. Humans naturally see change as risky because it is risky, just as mutations in genes are mostly destructive. You would not want to go to work were everything changed every week! The phone system, the office assignments, who reports to who, and the whole set of job expectations.

You overcome the logical resistance to change by proving that a new approach actually works, usually on a small scale.