Strategy isn't made more ambitious by ignoring the conditions in which it's gotta work. It's made less useful.
A credible product strategy begins with what users currently do, what your organization can currently support, where incentives conflict, which constraints are durable, and what remains uncertain. It also distinguishes the desired destination from the theory of how change will happen.
I'm not arguing for timid planning. Reality can reveal leverage as easily as limitation. A workaround may expose unmet demand. Repeated detachment from a shared system may reveal a missing capability. A support pattern may show that the organization is transferring its coordination burden to customers.
The strategic task is to form a view of the world that can be tested through product decisions. When evidence changes, the strategy should learn rather than defend its original language.
I didn't come up with this concept. Read more in Good Strategy, Bad Strategy or Hypothesis-Driven Development.