Semantic model similarity score

Compare every semantic model in your tenant and get a similarity score for each pair — the fastest way to find duplicates and near-duplicates worth consolidating.

Feature facts

Available in

—Free✓Enterprise✓Automation

Where in the app

Model similarity action in the top toolbar (after a scan) — see UI reference

Prerequisites

Desktop path — Measure Killer desktop app with a paid Enterprise license, or request a free trial
Automation path — an MK Automation subscription, run via the Notebook sample we provide on setup
A completed tenant or workspace scan — similarity is computed from scanned model metadata

Permissions required

Either: Fabric Administrator (or Power Platform Administrator) — similarity scored across every model in the tenant
Or: Workspace Contributor — similarity limited to the workspaces you have access to (runs in Limited Tenant Analysis mode)

Limited tenant analysis

Supported

Free trial

Available (with restrictions)Similarity view is fully visible during the trial. Exports (Excel, JSON) and automatic kill via XMLA are disabled until a paid licence is applied.

What you get

TODO: 2–3 sentences — describe how the score is computed (table / column / measure overlap), what counts as “duplicate”, the sort/filter UI.

Run the analysis

TODO: step-by-step — run a scan, open the similarity view, set a threshold, drill into a pair to see the diff.

Common workflows

TODO: 2–3 workflows — e.g. “Find consolidation candidates before a model-governance review”, “Spot copy-paste-and-edit drift across departments”.

Report similarity

TODO: companion view for Power BI report similarity — same score, applied to report metadata. Note that this is a separate compare row but shares the UI.

What to do with the findings

TODO: exports, planning a consolidation, who to escalate to.

TODO: link to /docs/how-tos/run-tenant-scan, /docs/how-tos/end-to-end-lineage.