Public Trust Signal Checklist

It applies broad text, URL, and link patterns to captured pages: HTTPS URL schemes, selected HTTP resource references, contact details, About and team wording, privacy, terms, cookie and refund references, trust and social-proof terms, company-detail phrases, selected review or payment providers, and possible CSP or referrer meta text.

No. It does not authenticate the organization, verify claims, test certificates or TLS, scan vulnerabilities, inspect transactions, evaluate customer service, check external reputation, or measure visitor perception. The score is a weighted pattern checklist.

No. Policy-link detection cannot determine which laws apply or whether a notice, consent flow, contract, refund practice, or data process is accurate and compliant. Review actual practices, users, transactions, technologies, and jurisdictions with qualified owners or advisers.

No. Broad terms such as customer, trusted, review, secure, Stripe, Google Reviews, or Trustpilot can match unrelated text. The crawl does not verify source, consent, ownership, certification, rating, partnership, payment acceptance, recency, or placement.

The crawler may not reach every page, raw HTML can be stripped from saved results, visual content and interactive widgets can be missed, and the configured wording is limited. Some signals are also irrelevant to particular organizations. Confirm each flag in context before making changes.

Treat them as review prompts. Verify the live page and business facts, identify the audience question each item should answer, publish only accurate and permissioned evidence, and use dedicated security, legal, reputation, analytics, and user-research methods for conclusions.