The problem
- Quality review is inconsistent: it depends on which editor is on shift and how much time they have.
- Editors cannot see patterns — which desks, formats, or source types repeatedly miss basic checks.
- New AI tools add output but not accountability, and nobody has time to verify at volume.
How MediaMeter helps
- Every draft gets the same seven-dimension assessment, weighted to its content type.
- Findings are ranked so editors see the three things that matter, not a hundred flags.
- Desk-level reporting shows where quality is improving and where it is slipping.
What changes for an editor
The sidebar appears in the CMS draft editor. It shows the current score, the dimensions that need attention, and inline evidence. The editor can accept a suggestion, dismiss it with a reason, or escalate to a second reader. Nothing is blocked.
A national desk uses the assessment to cut its pre-publication read time by focusing second reads on the 18 percent of drafts flagged for evidence or fairness. This is a design target, not a measured result.
Rollout
A pilot starts with one desk and one content type. We calibrate against that desk’s own standards, review disagreements weekly, and expand only when editors say the signal is trustworthy.