There isn’t a ‘best’ way. It depends on what information you are given and the question requirements set. You might have to carry out an NPV (like in FM) or to interpretation one, drawing conclusions and giving recommendation.
The only general point I would make is that in NPV calculations usually all amounts (even sometimes the initial investment) are estimates: future cash flows, timing, duration of project, scrap values discount rate. Therefore, it is often important to be sceptical about these assumptions and the results they give. Sensitivity analysis might be important depending on the question requirements.