This policy describes the workflow StayFront can apply today. It does not claim an expert panel, separate newsroom or professional review process that does not exist.
Source selection
We prefer current primary sources for factual claims about platform behavior, search systems, business profiles and technical implementation. Examples include official documentation from OpenAI, Google, Microsoft, Schema.org and the product being discussed.
Secondary research may be used when primary documentation does not answer the question. When it is used, the article should name the publisher, date and method where available, and should not present correlation as an official platform rule.
Claim verification and evidence labels
Claims that can change are checked against current sources during a substantive review. Citations are placed near the claim they support, followed by a readable source list.
We distinguish documented behavior from observation, supportive reasoning and unknowns. We do not publish private model internals, fixed AI ranking factors or placement guarantees as established facts without direct evidence.
Search and competitive research
Search-result research is used to understand the task behind a query, the language people use and the information current results cover. It is not used to copy a competitor's structure, wording, examples or claims.
StayFront does not publish invented keyword volume or traffic forecasts. When reliable first-party search data is unavailable, decisions are described through observed intent, result patterns, source quality and content gaps.
AI assistance and human review
AI tools may assist with research organization, drafting, rewriting, consistency checks and code implementation. AI output is not treated as a source and does not replace verification.
The publishing workflow requires a human review of the final page for accuracy, useful scope, source fit, conflicts, unsupported claims and business capability statements before the page is treated as approved content. Content that cannot complete that review remains draft, deferred or noindex.
Authorship
Resources use the organization-level byline StayFront.AI while verified, approved named-author information is unavailable. The byline does not imply a separate newsroom, editorial board, industry-expert review panel, medical reviewer or fact-check department.
A named author or reviewer will only be added when identity, role, consent and relevant qualifications are verified.
Published and reviewed dates
Published indicates when an article first entered the public resource library. Reviewed indicates a substantive factual or editorial review, not an automated build or cosmetic deployment.
A new date should not be added merely to make an article appear fresh. Platform behavior pages should be reviewed when official documentation or the described product experience materially changes.
Examples, observations and data
Illustrative examples are labeled or written as examples. They are not customer results, case studies or measured benchmarks. Observations from a manual test are described as observations and include enough context to understand their limits.
Statistics, benchmark claims and customer outcomes require a traceable source or a documented StayFront research method and real dataset. We do not create sample numbers that could be mistaken for measured results.
Commercial conflicts, affiliates and sponsorships
StayFront.AI sells AI visibility services, so content about AI visibility tools and methods can involve a commercial conflict. Relevant pages should state that relationship and should not present StayFront as objectively best without documented comparison evidence.
The current resource library does not rely on affiliate rankings or sponsored placement. If an article later includes an affiliate relationship, paid inclusion or sponsorship, that relationship must be disclosed clearly on the affected page.
Research and methodology
The product Methodology page explains the current measurement process. A future research report must separately document its question set, platforms, locations, time period, sample, exclusions, limitations and analysis before findings are published.
No case study, benchmark or industry ranking should be described as StayFront research until a real dataset and reviewable method exist.
Corrections
Readers can request a correction through the StayFront.AI contact form. Include the page URL, the statement in question and a supporting source when possible.
Material factual errors should be corrected promptly after verification. A correction that changes the article's conclusion or important guidance should be noted on the page. Minor grammar, formatting and link repairs may be made without a correction note.