Australian Agent-Readiness Census
The Australian Agent-Readiness Census
Whether Australian software vendors have done the work that lets an AI crawler read them at all. 54 Australian domains carry a comparable score. 8 categories have enough peers to report. Every figure below is an aggregate, and no business is named against its own score.
What we can report
| Category | Peers | Median | Middle half | Range |
|---|---|---|---|---|
| AML compliance software | 9 | 2/10 | 1 to 3 | 1 to 5 |
| tradie software | 7 | 1/10 | 1 to 3 | 0 to 4 |
| NDIS SIL software | 7 | 2/10 | 1 to 3.5 | 0 to 5 |
| allied health practice management software | 6 | 1/10 | 1 to 1 | 0 to 2 |
| childcare management software | 6 | 1.5/10 | 1 to 2 | 1 to 3 |
| conveyancing software | 5 | 1/10 | 1 to 1 | 1 to 3 |
| legal practice management software | 5 | 1/10 | 0 to 1 | 0 to 1 |
| property management software | 5 | 2/10 | 2 to 2 | 1 to 2 |
What we cannot report yet
These categories are below the 5-peer threshold, so no figure is given for them. They are listed because a census that hides its own gaps reads as complete when it is not. A category with nothing scanned is not a category that scores zero: those are different sentences and only one of them is an accusation.
- dental practice management software4 of 5 peers scanned
How this census is built
Each domain gets one deterministic scan of 17 agent-readiness checks. The score is the count of checks passed over the checks we could determine, on a 0 to 10 scale. The score is the count and nothing else. There is no weighting behind it. A check we could not complete is left out of the score rather than counted against you. A site we could not reach is not a site that failed.
A category is reported once at least 5 of its peers carry a comparable score. That is the same threshold we already apply before showing a single business its own position, deliberately: a lower bar to publish than to display would be indefensible the first time somebody asked why.
Quartiles are the R type 7 definition, linear interpolation on n-1, which is what a spreadsheet PERCENTILE() returns. A reader checking our arithmetic gets our number back.
8 audited Australian domains have a category but no comparable score yet, so category counts move as the backfill runs.
What these numbers do not say
There is no confidence interval here, by design. This is a census of one deterministic scan per domain, not a sample of a larger population. An interval would imply a sampling process that did not happen.
This measures readability, not visibility. Passing these checks makes your site readable to AI crawlers. On its own that is not evidence of more citations, and we do not sell it as one. We measure whether AI assistants cite a brand separately, on a different axis. The two are correlated observations on the same set of Australian businesses. Neither of them establishes that one causes the other, and we will not claim it does.
It is not revenue and it is not quality. A vendor scoring higher has implemented more of the checks. Nothing follows from that about how good their product is or how well they sell it.
This is the audited set, not the market. These are the Australian domains we have scanned. They are not a random draw from the Australian software market, so read them as a floor on what is out there rather than an estimate of it.
Individual results are never published. If you want to know where a specific Australian business sits, that goes to that business privately and to nobody else.
Check your own site free