In strategy, there is seldom one right answer. But framing appropriately the questions for analysis is vital. Ask the wrong questions (or the questions wrongly) and you will almost certainly get a wrong answer.
Over time, particular methods of analysis become entrenched. That does not always make them the most appropriate. Many commonly-used methods simply are not fit for the purposes to which they are applied.
The provenance of data will often determine its validity and the confidence the analyst can place in it. Also, for contested measures, gaining data from a variety of sources improves the dependability of analysis.
The presentation of the results of analysis will often determine its impact. Error statistics and confidence levels matter, as do the values of other options and the reliability of data sources.