Peer benchmarking

The current position

No board pack contains a peer comparison. The benchmark dataset ships empty, with no source, no sample size, no methodology and no date, and an empty dataset disables benchmarking entirely.

Why it ships disabled

A benchmark is a claim about a population. Publishing "your security posture is below the industry average" requires knowing which organisations are in that average, how many, when they were measured, and by what method.

Without those, the sentence is a rhetorical device with a number in it. That is the correct default until a real peer dataset with a stated methodology exists.

What would have to be true for it to appear

The engine emits a comparison only when the dataset declares:

• a named source
• a sample size greater than zero
• an as-of date
• a stated methodology

If any is absent, no comparison is produced. There is no partial or best-effort mode.

What you get instead

Your own position, measured against a published methodology you can check, and — where a previous assessment exists — movement against your own prior measurement. See What changed since the last assessment.

Comparison against yourself is a stronger governance statement than comparison against an unnamed cohort, and it is one GovIntel can actually evidence.

Status

This page describes current behaviour. Benchmarking is not a planned feature with a date; it is a capability gated behind a dataset that does not exist.