Top-down vs bottom-up approach

In implementing a data strategy, we need the right balance between the top-down and the bottom-up approach to try and have the advantages of both, mitigating the disadvantages as much as possible.

We can imagine Plato and Aristotle discussing the ideal architecture for their data infrastructure.

Plato proclaims the existence of an ideal world behind our perception, where we recognize things because our soul remembers them, coming from this ideal world itself.

Aristotle claims we recognize something because we derive common traits in different but similar objects in the world. A top-down vs a bottom-up approach to cognition that we can apply to data projects as well.

Top-down means senior management deciding on a strategy and implementing processes, systems and KPI's to follow up on this strategy. Advantage is to have a clear direction and to go fast, disadvantage is that change is difficult and without proper explanation can entice a lot of resistance in the organization.

Bottom-up means running workshops with people on the front lines of the business. With the people servicing the customers directly and asking for their input to the strategy and relevant KPI's. Advantage is to create buy-in, disadvantage is that this takes time, and it might prove difficult to find consensus between the different people involved in the workshops to land on a clear overarching strategy.

What are the advantages and disadvantages of top-down and bottom-up approaches that you have seen in your organization?

If you had to choose one, would you choose bottom-up or top-down?