1.2 Projects as decision systems

A useful way to understand a project is as a chain of connected decisions:

1. Problem decision: What issue deserves investigation?
2. Evidence decision: Which data and secondary sources are relevant?
3. Measurement decision: How will concepts such as demand, value, risk, retention, or success be represented?
4. Method decision: Which analytical approach fits the question and evidence?
5. Interpretation decision: What conclusions are supported, and what remains uncertain?
6. Action decision: What should the client consider doing?

An error early in this chain can travel through the entire project. For example, if customer activity is measured using transactions rather than unique customers, later claims about customer growth may be misleading. If a projection uses a population denominator that does not match the client’s service area, the final capacity recommendation may be inaccurate even when the arithmetic is correct.