A fixed rubric
Every criterion has a written test, and none of them names a company. The same question is asked of every report, so two reports can be compared at all.
Emperor · Middle East
What makes an annual report good, argued from evidence.
Ten sections. Fifty-three criteria, each with a written test. One hundred points, every one of them traceable to a page.
Invite-only. Ask your administrator for access.
Ten sectionshover an axis
A shared peer set
Every report scored becomes a benchmark for everyone. Comparisons come from reports that were actually read, not from impressions of the market.
A human confirms it
A model proposes the scores. Nothing enters the database until someone has been through all fifty-three and confirmed them, and every override is recorded against a name.
Two sections are capped.
Highlights and the investment case must appear as a discrete, signposted block. A report can satisfy every criterion in them and still lose the difference for scattering it across ninety pages — because a reader who cannot find it has not been told.
See how the caps work