A logframe typically begins to take shape by defining the ultimate goal, followed by the purpose of the project, then the outputs needed to achieve the goal, and finally, the activities and inputs needed to achieve the outputs. Only one goal and purpose should be stated for each project or the project will risk being unfocused. Normally, however, there are multiple activities and outputs in a project, and these are reflected in the logframe.
One of the ‘tricks’ of the logframe is to know how to test its underlying logic. This is done by reading the logframe from bottom to top to analyze the coherence of its arguments (see figure 2). For example, the linkages between the components of the matrix would be read as follows: if activities (as listed in cell A4) are implemented, and the relevant assumptions are valid (cell D4), then the project will achieve the outputs (cell A3). If outputs are achieved and the related assumption remain valid (cell D3), the project will achieve its purpose (cell A2). If the purpose is achieved and the related assumption holds (cell D2), then the overall goal is achieved (A1). The middle columns (B and C) show what and how to measure the achievement of the summary in each level. These are indicators and means of verification respectively. The logframe will fail on its logic, if at any point the statements of inputs, activities, outputs, purpose or goal are not clearly related, or if essential information is missing.