Four pillars of a machine-surface score
A useful public diagnostic needs a bounded vocabulary. Without one, every scan invents its own language for "good", and scores cannot be compared, disputed or re-run against the same evidence. Brittek's Machine Surface Inspector groups deterministic public evidence into four equal pillars: Findability, Quotability, Structural integrity and Trust. Each is a finite question set with a point budget — not a brand metaphor.
- Published
- 5 Sept 2026
- Reading
- 5 min
- Published by
- Brittek Digital
The pillars
Findability asks whether the page can be reached and identified: crawl access, sitemap validity, canonical declarations and redirect behaviour.
Quotability asks whether the page yields extractable answers: content depth, heading structure, freshness and answer-shaped patterns a retrieval system can reuse.
Structural integrity asks whether document meaning survives machine reading: JSON-LD accuracy, Open Graph coverage and semantic document architecture.
Trust asks whether authority signals are present for a retrieval system: HTTPS, identity, contact, policy and related signals.
Score as summary, not product
As the deterministic diagnostics note argues, a number orients; it does not replace the evidence. The product is the retained input, the explicit rules, the recoverable points and the remediation that would change the result. Unavailable evidence must not silently pass.
Why this compounds
- Clients and operators share one vocabulary for remediation.
- Regression tests can re-score the same fixture.
- Journal, tools and machine files stay aligned on what "machine-readable" means in practice.