Criteria are set before conclusions. Every claim is traceable to evidence. Editorial interpretation is labeled as such. Corrections are reviewed within 14 days. AI assistance is disclosed on any page where it was used.
Editorial standards
Criteria before conclusions
Brightfield Research establishes evaluation criteria before assessing any subject. This is the single most important structural rule in our editorial process. Criteria designed after assessment has begun are criteria shaped by the conclusions already forming in the researcher's mind. That sequence produces outputs that look like structured analysis but function as post-hoc justification. We do not permit it.
Every research initiative begins with a documented criteria framework. That framework specifies what characteristics the research will evaluate, why those characteristics are relevant to the decision context the research addresses, and how the presence or absence of evidence for each characteristic will be treated. The framework is reviewed before the evidence collection phase begins.
It is not revised after individual subjects have been assessed without editorial sign-off and a documented rationale for the revision.
This requirement is not a procedural formality. It exists because research consumers have a legitimate expectation that the framework applied to a subject was not designed around that subject's particular strengths. When criteria are set in advance and applied consistently across all subjects in scope, the conclusions that result are structurally more defensible, even if they are not more comfortable for the subjects they evaluate.
Claims traceable to evidence
Every factual claim in a Brightfield Research publication must be traceable to a documented source. This traceability requirement applies without exception to claims about a subject's attributes, capabilities, history, commercial terms, certifications, qualifications, market position, or operational characteristics.
Traceability does not mean that every claim carries an inline citation in the body of the published text. It means that a source note for the claim exists, that the source is accessible or was accessible at the time of research, and that the editorial record contains a record of what the source stated and when it was reviewed.
Readers can request source notes for any specific claim by contacting the editorial desk. Source notes for all published outputs are available on request.
Claims that cannot be traced to a documented source are either omitted or labeled as editorial inference. Editorial inference is a legitimate category of content in Brightfield Research outputs, but it must be identified as such. Readers are entitled to know whether a statement in a published piece is a documented fact, an editorial judgment, or a reported claim from a third-party source that has not been independently verified by Brightfield.
Where a claim originates from a subject's own promotional materials, that provenance is noted. Claims from promotional materials carry reduced interpretive weight in our evidence framework and are not treated as independently verified facts about the subject's actual attributes or performance.
Editorial interpretation labeled
Research involves judgment. Brightfield Research does not pretend otherwise. Selecting which evidence to weight most heavily, which source gaps to treat as significant, how to characterize the competitive position of a category, and what a set of documented facts implies for a decision-maker all require editorial interpretation. We do not hide that interpretation behind language that implies machine-like objectivity.
Editorial interpretation in Brightfield Research outputs is identified explicitly. Statements of documented fact are presented as statements of documented fact. Statements of editorial judgment are presented as editorial judgment, typically with language that signals the interpretive nature of the statement: "Brightfield's assessment," "based on available evidence," "the editorial team's reading of this evidence is," or similar formulations.
The distinction matters. Readers who disagree with an editorial judgment can challenge the interpretation. Readers who believe a documented fact is wrong can submit evidence of the error through the corrections process. Both types of challenge are legitimate and both are addressed through different pathways.
Voice and tone
Active and specific
Brightfield Research publishes in active voice. Passive constructions that obscure who is making a claim, who performed an action, or who holds a position are not permitted in final published text. "The company has been recognized for its service quality" is not acceptable. "The company received the Stevie Award for Customer Service Excellence in the manufacturing sector in 2024" is acceptable, if traceable to a documented source.
Specificity is the companion requirement to active voice. Vague claims that sound informative but communicate nothing verifiable are excluded from published text. "A wide range of clients," "significant experience," "comprehensive solutions," and similar formulations that offer no measurable content are editorial filler. They appear frequently in promotional content because they are difficult to disprove. They do not appear in Brightfield Research publications because they do not help decision-makers make better decisions.
Defensible
Every statement in a Brightfield Research publication should be defensible by a named member of the editorial team who can point to the evidence that supports it and explain why that evidence warrants the statement made. This is the practical test applied during editorial review. If an editor cannot defend a statement in response to a reader challenge, the statement is either revised or removed before publication.
Defensibility is not the same as certainty. Many statements in research outputs are defensible precisely because they acknowledge the limits of available evidence. "The company does not publish pricing information publicly and the editorial team was unable to obtain verified pricing data" is a defensible statement that contributes to a reader's understanding.
"The company offers competitive pricing" is not a defensible statement because it makes a comparative claim without specifying the comparison or the evidence on which the comparison rests.
No superlatives without evidence
Superlative language is among the most common forms of unverifiable content in research adjacent material. Claims that a subject is "the leading provider," "the most trusted," "the industry standard," or "the best" in any category require, at minimum, a documented source for the superlative that is independent of the subject making the claim, a defined scope within which the superlative applies, and a defined time period.
Brightfield Research does not use superlative language without meeting all three of these requirements. Where a subject holds a position that warrants superlative language, the evidence for that position is cited directly. Where the evidence does not support a superlative claim, the claim is not made, regardless of how frequently the claim appears in the subject's promotional materials or how widely it has been repeated in third-party coverage that did not independently verify it.
Content structure requirements
Brightfield Research publications follow a consistent structure that reflects the organization of the evidence. Structure serves the reader's ability to evaluate the research, not a presentation convention.
Required elements for all published outputs
Research conclusion stated at the top in plain terms. Decision-makers who need only the headline finding should be able to locate it immediately.
A documented statement of what category the research covers, who the intended decision-maker is, and what decision context the research is designed to support. Scope limitations are stated here.
The criteria applied in the research, published in a form that allows a reader to evaluate whether the criteria are appropriate for the stated decision context and whether they were applied consistently across all subjects in scope.
The evidence reviewed, organized by subject or by criterion depending on the format, with source type identified and interpretive weight noted where relevant.
An explicit statement of what the research does not cover, where the evidence was insufficient, what assumptions were made in the absence of documentation, and what has changed or may change.
A documented record of the sources consulted in producing the output, organized to allow a reader to verify the primary claims in the text.
A statement of any commercial relationships relevant to the output. Where no commercial relationships exist, this is stated explicitly.
Contact information for the corrections process and a statement of how corrections are reviewed and applied.
Publication date and review date
All published outputs carry a publication date. Where a published output has been reviewed and confirmed current after initial publication, the last-reviewed date is shown alongside the publication date. Where a published output has been materially updated, the update is documented in the corrections and updates log.
Corrections policy
How to submit a correction
Corrections are submitted by email to [email protected]. A complete submission includes:
Link to the page containing the claim being challenged.
Quoted accurately from the published text, with the section or paragraph identified.
What you believe the correct claim is, in plain terms.
The source of the contradicting evidence, with sufficient detail for the editorial team to verify it independently.
In case the editorial team needs clarification on the submission.
Brightfield Research accepts correction submissions from any reader, including subjects of research, competitors of subjects of research, and parties with disclosed commercial interests in the outcome. The source of a correction submission does not determine its merit. The quality of the submitted evidence determines its merit. All submissions are reviewed against the same standard regardless of who submits them.
What constitutes a correctable error
A correctable error is a factual statement in a published Brightfield Research output that is demonstrably inaccurate based on documented evidence that was available at the time of publication or that has become available since publication. Correctable errors include:
- Factual claims about a subject's attributes, history, or certifications that are contradicted by the subject's own verified documentation or by independent third-party records
- Source attributions that misrepresent what the cited source actually states
- Dates, figures, or statistics that are demonstrably incorrect
- Statements of scope or coverage that do not accurately reflect the research conducted
- Material omissions that cause a published conclusion to be misleading, where the omitted information was publicly available at the time of publication
Corrections are not accepted for editorial judgments that the submitter disagrees with, where the disagreement is a difference of interpretation rather than a documented factual error. A subject who believes Brightfield's assessment of their evidence quality is too harsh is not submitting a correctable error. A subject who can document that a specific factual claim about their certification history is wrong is submitting a correctable error.
Corrections are also not accepted for information that has changed after the research was published unless the change is material to a decision being made on the basis of the published research. Where changed circumstances are reported, the editorial team may add a note to the published output reflecting the change without treating the original publication as erroneous.
Turnaround
Brightfield Research reviews all correction submissions and responds to the submitter within 14 calendar days of receipt. Submissions received at periods of high editorial volume may require the full 14-day window. Where a correction involves complex evidence assessment or consultation with a reviewer, the editorial team notifies the submitter of the extended timeline.
If a correction is accepted, the published page is updated within 7 days of the acceptance decision. The update is noted on the page with a date and a brief description of what was changed. The original text is not suppressed. Where editorial process permits, the original text is preserved alongside the corrected text in a way that allows readers to understand what changed and why.
Declined corrections
If a correction submission is declined, the editorial team provides a written explanation to the submitter. The explanation identifies what evidence was reviewed, why the submitted evidence did not, in the editorial team's assessment, support the proposed change, and what evidence would be needed to support reconsideration. Corrections are not declined to protect editorial convenience, commercial relationships, or the reputation of the publication. They are declined only when the submitted evidence does not, after review, support the proposed change.
A submitter who receives a declined correction and believes the editorial team's assessment is wrong may escalate the submission by providing additional evidence. There is no formal appeals body. Additional evidence is reviewed by a different member of the editorial team from the one who conducted the initial review.
AI assistance disclosure
Brightfield Research uses AI tools in editorial workflows. This section describes how AI tools are used, what limits apply, and how that usage is disclosed to readers.
Permitted uses
AI assistance is used in Brightfield Research editorial workflows for the following purposes:
- Drafting. AI tools may be used to produce initial drafts of published content based on research notes and evidence compiled by the editorial team. AI-produced drafts are not published without review, revision, and editorial sign-off by a named member of the team.
- Summarization. AI tools may be used to summarize source documents, extract relevant passages from lengthy evidence sets, and organize research notes into a structured form suitable for editorial review.
- Preliminary organization. AI tools may be used to produce a preliminary organizational structure for a research output based on the criteria framework and evidence set assembled by the research team.
What AI tools do not determine
AI tools do not determine the criteria framework applied in any research output. They do not determine which subjects fall within scope of a research initiative. They do not make editorial judgments about evidence quality or the conclusions that evidence supports. They do not verify sources, conduct independent research beyond what the editorial team has assembled, or produce claims without editorial review of the underlying evidence.
Same evidence standard applies
Content produced with AI assistance is subject to the same evidence standards as content produced entirely by human writers. AI-produced text that contains claims without traceable sources is revised or removed during editorial review. AI-produced text that uses superlative language without evidence is revised or removed. AI-produced text that obscures the interpretive nature of editorial judgments is revised to make the interpretation explicit.
The workflow that produced a draft does not change the evidence standard the draft must meet before publication.
Disclosure on published pages
Where AI assistance was used in producing a published output, this is disclosed on the published page. The disclosure appears in the production notes section of the output and identifies the nature of the AI assistance (drafting, summarization, or both). Where AI assistance was not used in producing an output, no disclosure appears because no disclosure is required.
Absence of disclosure does not imply that AI tools are never used. It means AI tools were not used for that specific output.
Prohibited content
The following types of content are prohibited from publication under any circumstances in any Brightfield Research output, regardless of research format, category, or commercial or editorial pressure.
Numerical claims — percentages, market share figures, survey results — that are not traceable to a real documented source. If the number cannot be sourced, it does not appear.
References to sources that do not exist. If a source is cited, it exists. If the editorial team cannot verify a source during the research process, the citation is removed.
Statements attributing to Brightfield Research a history, credential, or experience that it does not hold. Brightfield Research was founded in 2026. Any publication suggesting an earlier history is erroneous.
Scores, ratings, rankings, or comparative assessments produced by a methodology that is not publicly documented. Where assessments are published, the criteria and weighting methodology are published alongside them.
Editorial conclusions influenced by commercial payment. No organization can pay for a favorable conclusion, a higher position in a comparative assessment, or inclusion in a research output. Where commercial relationships exist, they are disclosed. They do not influence conclusions.
Credentials attributed to named reviewers that those reviewers do not hold, or reviewer profiles for individuals whose identity has not been verified. Reviewer profiles are not activated until identity verification, credential review, and conflict disclosure are complete.
Frequently asked questions
Brightfield Research will never publish fabricated statistics or invented data presented as research findings; invented citations or references to sources that do not exist; claimed history, credentials, or qualifications that the organization does not hold; undisclosed scoring methodologies used to produce rankings or comparative assessments; pay-to-rank placements or any editorial conclusion influenced by commercial payment; ghost credentials attributed to reviewers who have not verified their identity and scope; or superlative claims about a subject that are not traceable to specific documented evidence.
Corrections can be submitted by email to [email protected]. A submission should include the specific claim being challenged, the URL and section where it appears, and documented evidence supporting the correction. The editorial team reviews all submissions and responds within 14 calendar days. If a correction is accepted, the published page is updated, the correction is logged with a date, and the nature of the change is noted. If a correction is declined, the submitter receives a written explanation of the editorial reasoning. Brightfield does not decline corrections to avoid reputational inconvenience. Corrections are declined only when the submitted evidence does not, after review, support the proposed change.
Yes. Brightfield Research uses AI assistance in editorial workflows for tasks including drafting, summarization, and preliminary organization of research notes. AI assistance is disclosed on any page where it was used in production. All AI-assisted outputs are reviewed by a named member of the editorial team before publication. The same evidence standards apply regardless of whether a draft originated with a human writer or an AI tool. AI outputs are not published without editorial review and are never used as a substitute for documented evidence.
Updates and versioning
This editorial policy is maintained as a living document. The Brightfield Research editorial team reviews it on a scheduled basis and updates it when significant changes to editorial practice, legal context, or platform capabilities require revision.
The current version of this policy is version 1.0, effective June 2026. The policy was established at the founding of Brightfield Research as the governing document for all editorial decisions. Prior to version 1.0, this policy did not exist in published form because Brightfield Research did not publish in published form. Version 1.0 reflects the standards applied from the founding of the publication.
When this policy is updated, the version number increments, the effective date is updated, and the changes are noted. Where a change to editorial policy has a material effect on previously published research outputs, those outputs are reviewed and updated where necessary. Readers who have questions about how a specific policy change affects a specific published output can direct questions to [email protected].
Periodic review does not mean periodic weakening. The intent of scheduled review is to strengthen the policy where practice has revealed gaps, to extend it to cover new editorial situations that were not anticipated at the time of the prior version, and to ensure it remains consistent with the legal and professional environment in which Brightfield Research operates.
Changes that weaken the evidence standards or the prohibited content rules require a documented rationale and sign-off from the publisher.