1.4 Integrated project evidence

Many applied projects require several kinds of evidence. Depending on the problem, a project may include:

• a profile or segmentation of important populations, customers, services, products, or regions;
• an appropriate descriptive, statistical, or machine-learning analysis;
• a forecast or conditional projection;
• relevant secondary research;
• problems or opportunities identified primarily from the organization’s data; and
• realistic recommendations that can be traced to the analysis.

These elements should form one argument. They should not appear as independent demonstrations joined only by a table of contents. Segmentation should clarify which groups matter. Modelling should answer a defined question. Forecasting should show the future consequence of a pattern or assumption. Secondary research should improve interpretation. Recommendations should respond to the problems established by the evidence.