The stack rank is the argument

You can't stack-rank brands within a vertical without committing to claims. The ordering is the argument — evidence on the table, reasoning traceable, disagreement testable. This is the opposite of the average vendor deck, which evades commitment through parallelism ("here are the ten leaders, everyone's strong in something").

Parallelism is an epistemic dodge. Ranking is an epistemic bet. The analyst has to decide, and the reasoning has to show. When the rubric is public and the rank is defensible, the report operates as an instrument rather than a pitch. When it isn't, the rank is noise.

What this lets stakeholders do: argue with the analyst on merits rather than vibes. Trace a disagreement to a specific line item in the rubric. Update their own view when the evidence is tighter than their prior.

What's still open: how do you rank within a vertical where the participants won't sit still long enough to be compared?