Data should be collected as set out during preparation, from the functions concerned, in the formats agreed whether online, offline, by questionnaire, interview or desk research, and it should be sound in quality and cover each key success factor together with the strategy and the way that strategy is put into practice. Where an assessment is large, run a pilot first, check it for unusual patterns and outliers that might mask important factors or point to misunderstanding, then adjust before collecting at full scale, picking benchmarking classes that fit what the assessment is meant to achieve.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.