Public-interface research protocol.

Purpose

Public-interface research can reveal how a business frames an offer, guides a decision and opens a route to contact. It cannot reveal what happens inside the business after a visitor acts. The method begins by keeping those two kinds of knowledge separate.

Unit of observation

The unit is a publicly accessible journey: a page, search path, form, booking route, appointment flow, contact path or related decision point visible without an account or a transaction.

Each observation records the source URL, access date, relevant interface state, visible element, interpretation and a limitation. A source is not treated as evidence beyond what was visible in that state.

What is coded

• Orientation: what the site says it is for and who it serves.
• Choice: the entry points available to visitors with different levels of certainty.
• Context: the information requested before a next step.
• Handoff: the person, role, channel or team named for the next interaction.
• Continuity: the visible explanation of what follows an inquiry, booking or request.
• Boundary: disclosed terms, eligibility conditions, privacy limits or trade-offs adjacent to a decision.

What is not inferred

The research does not infer conversion rate, revenue, market share, customer satisfaction, service quality, operational capacity, legal compliance, availability, response time, accessibility compliance or security from a public page alone. It does not rank businesses or describe them as “best” based on website review.

Ethical and practical limits

No form is submitted, no account is created and no booking, payment or appointment is attempted solely for research. Region-specific, session-specific, blocked or redirected pages are recorded as limitations rather than negative findings. Screenshots, when retained, document a page state; they do not license another brand’s text, imagery or interface.

How findings become recommendations

Observations are translated into general design questions rather than copied implementations. A public booking path may suggest that a visitor benefits from process clarity; it does not mean a small service studio should reproduce a global booking engine.

Recommendations for a client website should be tied to the client’s actual service, audience, operational capacity and evidence—not to a competitor’s visual surface.

Reproducibility

Every published study should state its access date, sample definition, observation limits and source register. A reader should be able to distinguish the source material from the author’s interpretation and the interpretation from any practical recommendation.