Methodology
Built for business websites.
PARSE is our proprietary methodology for measuring how well AI agents can reach, read and act on your site. We publish how the score works. The individual checks stay private. Fixed rules set the score.
What we check
Every audit tests your pages against a library of specific checks in three pillars.
Can agents get in?
Whether your site’s front door is open to AI agents at all: what your robots rules allow, how your pages load, and what a machine can actually retrieve.
Can agents act?
Whether an agent trying to contact you, find your services or take the next step can find the signposts it needs.
Do agents understand?
Whether what your pages say is structured so a machine reads the same meaning a person does.
Built for business websites
The checks fit the sites most businesses run: services, shops, practices and agencies. Checks that do not apply to your site are not counted.
We read your site, not just your home page. A free audit reads your home page plus up to 20 more pages, or your whole site if it has fewer. A detailed audit reads your home page plus up to 150 more pages, or your whole site if it has fewer.
How the score works
Each check that fails is a finding, rated Critical, Major or Minor.
In each pillar, a Major finding counts twice as much as a Minor one, measured against the checks that apply to your site. The three pillar scores add up to your PARSE Score out of 100.
Critical findings cap the score. If AI agents cannot get into your site, or find no content when they do, the verdict is Digitally Anonymous whatever the number. Other Critical findings hold the score at 70 or below.
| Verdict | PARSE Score |
|---|---|
| Agentic Authority | 85 to 100 |
| Agentic Strain | 60 to 84 |
| Digitally Anonymous | 0 to 59 |
Same site, same score
No AI model sets your score. Fixed rules do. Pages are chosen the same way on every audit, and every finding shows the evidence we found.
What you get
Free audit: your PARSE Score, your verdict and your finding counts, on screen and by email.
Detailed audit: every finding ranked by severity, the pages affected, the evidence found, step-by-step remediation, plus the page-by-page breakdown, the Evidence Vault, and Excel and PDF downloads.