Editorial Standards

3D Chess Media, LLC develops and operates research, publishing, marketing, and data businesses within The 3D Chess Media Portfolio. Our properties cover different industries and may use different research methodologies, editorial processes, subject-matter reviewers, data sources, and commercial models. This page describes the general editorial principles we want those properties to follow. Individual websites may maintain more specific standards appropriate to their subject matter.

Our Editorial Objective

Our goal is to publish information that helps readers understand:

What the research found
How the research was conducted
What the data can reasonably support
Where uncertainty exists
What commercial relationships may be relevant
Where human judgment has been used
What limitations readers should consider

We do not believe credibility requires pretending that commercial interests, methodological choices, or uncertainty do not exist. We believe they should be made visible.

Separate Observation From Interpretation

One of our most important editorial principles is distinguishing what was observed from how that observation is interpreted. For example:

Observed result: Six of seven AI systems recommended Company A.

Interpretation: Company A demonstrated strong cross-platform recommendation visibility in the systems and prompts tested.

Those statements are related. They are not identical. The observed result is the underlying data. The interpretation is our explanation of what the data may mean. Whenever practical, readers should be able to tell the difference.

Do Not Turn AI Agreement Into Objective Truth

Several 3D Chess Media properties analyze recommendations made by multiple AI systems. If several systems recommend the same company, we may describe that as: AI consensus, Cross-platform agreement, Recommendation agreement, and Consensus among the systems tested. That does not automatically mean the company is objectively the best choice for every person. AI systems may share:

Search results
Training information
Publishers
Company information
Reviews
Research
Other underlying sources

Therefore: 6 of 7 AI systems recommended Company A, should not be rewritten as: Six independent experts proved Company A is the best company. Our editorial language should reflect what the research actually measured.

Preserve Disagreement

Not every study needs a clear winner. If AI systems disagree, our reporting should show that disagreement. If: ChatGPT recommends Company A, Gemini recommends Company B, Claude recommends Company C, that may reveal something important about the market. We should not remove legitimate disagreement merely to produce a simpler headline. Disagreement can indicate:

Different source environments
Different platform behavior
Weak category consensus
Prompt sensitivity
Different interpretations of buyer intent
Differences in brand positioning

An uncertain result is still a result.

Commercial Relationships Should Not Rewrite Research

3D Chess Media operates commercial businesses. Portfolio properties may earn revenue from:

Affiliate programs
Advertising
Lead generation
Client services
Sponsorship
Research products
Data products
Referral relationships
Strategic partnerships
Subscriptions
Licensing

Those relationships can create potential conflicts. Our preferred editorial response is: disclose the relationship, and preserve the underlying result. If a commercial partner performs poorly in a defined research methodology, the data should be allowed to show that. If a company with no commercial relationship performs strongly, the data should be allowed to show that too.

A Partner Does Not Receive Extra Credit

If an underlying study finds: Commercial Partner A: recommended by 2 of 7 systems, and Non-Partner B: recommended by 6 of 7 systems, the commercial relationship should not change those observations. Affiliate compensation, advertising, sponsorship, or a client relationship should not manufacture additional AI recommendations. Likewise, a commercial relationship should not automatically remove a company from research if the methodology otherwise calls for its inclusion.

The 3D Chess Media Portfolio includes both research businesses and commercial service businesses. For example: LLM Authority Index focuses on AI search measurement, datasets, citation research, benchmarking, and competitive intelligence. CiteWorks Studio provides AI search strategy and implementation services. If either business appears in research concerning a market in which it participates, that relationship creates a potential conflict. Our preferred approach is to: Apply the same methodology., Preserve the result., and Disclose the ownership relationship..

We do not believe the best solution is to hide the relationship.

Human Review Should Add Context, Not Manufacture Outcomes

Some portfolio properties may use subject-matter reviewers. A reviewer may help with:

Factual accuracy
Industry terminology
Material omissions
Research interpretation
Consumer context
Methodological issues
Risk
Claims
Clarity

A reviewer should not change an observed AI response simply because they personally disagree with it. If an AI system recommended Company B, that remains the observed response even if the reviewer would personally choose Company A. The reviewer may explain why the recommendation deserves additional context. That is different from rewriting the research.

Subject-Matter Expertise

Different markets require different levels of specialist knowledge. Properties covering areas such as: Credit, Mortgages, Insurance, Aging in place, Home services, Business software, and Technology. may benefit from reviewers with relevant professional or industry experience. Where a reviewer is used, the applicable website should make clear:

Who the reviewer is
Why their experience is relevant
What they reviewed
What they did not review
Any material conflicts of interest

A reviewer title should not imply involvement beyond the work actually performed.

Editorial Independence From Advertisers

Advertising may appear on certain portfolio properties. An advertiser should not automatically receive:

Better research results
Higher rankings
Preferential recommendation treatment
Removal of unfavorable data
Guaranteed positive editorial commentary

Paid promotional content should be distinguishable from independent editorial or research content where the distinction is material.

Sources

Our properties may use information from sources including:

Government agencies
Company websites
Regulatory organizations
Academic research
Industry associations
Publishers
Professional organizations
Public databases
Product documentation
AI-generated responses
Original research
Proprietary datasets
Interviews
Subject-matter experts

The appropriate source depends on the claim. Whenever practical, important factual statements should be grounded in sources that are reasonably authoritative for the subject.

Primary Sources

For factual information about a company, product, policy, regulation, or government program, primary sources are often preferable when available. Examples may include:

Official company documentation
Government websites
Regulatory filings
Official pricing pages
Product documentation
Published terms
Public datasets
Original reports

A secondary source may still provide useful analysis or context. But when a definitive primary source exists, it should generally receive greater evidentiary weight.

Company Claims

Companies are often the best source for information about:

Their own products
Features
Pricing
Service areas
Leadership
Company history
Policies

But a company is not necessarily the best independent source for claims such as: We are the most trusted company in America. We provide the best service. Customers prefer us to every competitor. Those claims may require independent support if presented as objective facts. Our editorial process should distinguish between: the company says, and independent evidence shows.

Reviews and Reputation Data

When using: Customer reviews, Star ratings, Complaint information, Consumer feedback, and Community discussions. context matters. Review platforms can differ in:

Audience
Verification
Moderation
Collection methods
Sample size
Incentives
Fraud prevention

A rating from one platform should not automatically be treated as a universal measure of customer satisfaction. Where material, the source and limitations should be identified.

AI-Generated Information

AI systems can provide valuable research data. They can also produce:

Errors
Outdated information
Unsupported claims
Fabricated citations
Inconsistent answers

When AI responses are the subject of the research, the response itself is important evidence of what the system produced. But that does not mean every factual statement inside the AI response should automatically be accepted as true. For example: If our research asks which companies ChatGPT recommends, ChatGPT's answer is evidence of the recommendation. It is not automatically proof that every reason ChatGPT gave for the recommendation is factually correct.

Dates Matter

Research can become outdated. This is particularly true for:

AI recommendations
Pricing
Software features
Product availability
Interest rates
Insurance products
Company ownership
Regulations
Executive roles

Where timing materially affects interpretation, content should identify the relevant:

Research date
Collection period
Publication date
Update date

Updates

Portfolio properties may periodically update:

Research
Reviews
Comparisons
Rankings
Pricing
Company information
Methodologies
Disclosures

An updated page may incorporate new information without necessarily representing a completely new study. Where a material methodology changes, that change should be identified when it would affect comparisons with earlier results.

Corrections

We want material factual errors corrected. Examples might include:

Incorrect company names
Incorrect pricing
Incorrect ownership information
Incorrect professional credentials
Data-processing errors
Misidentified companies
Incorrect citations
Material calculation errors

A correction should address the error rather than quietly preserve information we know to be wrong.

Corrections Should Not Be Purchased

A company does not need to:

Become a client
Purchase advertising
Join an affiliate program
Sponsor research
Pay a fee

To request correction of a factual error. Likewise, refusing a commercial relationship should not prevent a legitimate correction.

Disagreement Is Not Automatically an Error

A company may dislike:

A ranking
A research conclusion
An AI recommendation result
An editorial interpretation
A comparison

That does not necessarily mean the publication contains a factual error. We may distinguish between: a factual correction, and a disagreement with the research result. If the underlying data was collected and reported accurately, we should not describe the result as erroneous merely because one participant disagrees with it.

Company Responses

Where appropriate, a company may be invited to:

Correct factual information
Provide additional context
Explain a position
Respond to material criticism
Clarify product details

Providing a company the opportunity to respond does not mean the company controls the final publication.

Rankings

If a portfolio property creates a ranking, the ranking should have a defined basis. Depending on the property, that basis may include:

AI recommendation frequency
Cross-platform agreement
Recommendation position
Product characteristics
Expert review
Pricing
Consumer suitability
Other defined factors

The publication should avoid implying that a ranking is purely objective if substantial editorial judgment is involved.

Data-Based Rankings

Where a ranking is directly calculated from a dataset, the methodology should be sufficiently clear to understand what is being measured. For example: If a ranking is based solely on how many AI systems recommended a company, that should be stated. It should not be presented as though we independently evaluated every product feature unless we actually did.

Editorial Rankings

Some content may include editorial judgment. Where human judgment materially influences a ranking or selection, we should avoid presenting the result as though it were generated entirely by objective data. There is nothing inherently wrong with editorial judgment. The important issue is describing the process accurately.

Reviews

A review may contain multiple forms of information:

Objective product facts
Research data
AI recommendation data
Consumer considerations
Reviewer interpretation
Editorial opinion

These should not be blurred together unnecessarily. The stronger the distinction, the easier it is for readers to understand what they are evaluating.

Comparisons

Comparison content should attempt to compare companies on criteria relevant to the question being asked. Those criteria may include:

Price
Features
Availability
Customer type
Use case
Support
Reputation
Technology
AI recommendation visibility
Other factors

The best company for one situation may not be the best company for another. We should avoid unnecessary universal claims when the evidence supports a more specific conclusion.

Pricing

Pricing changes frequently. When publishing pricing information, we may rely on:

Official company pricing
Public quotes
Published plans
Research observations
Estimated ranges

If a price is an estimate, range, starting price, or modeled figure, the distinction should be clear. Readers should verify current pricing directly with the provider before making a significant purchase decision.

Affiliate Links

Affiliate links may generate compensation for a 3D Chess Media property. When applicable, disclosure should explain that compensation may be received. The existence of an affiliate link should not be concealed in a way that would materially mislead a reader about the commercial relationship. For additional information, see:

Advertising & Affiliate Disclosure

Headlines

Headlines should communicate the substance of the underlying content without materially overstating what the research proves. For example, if a study measures AI recommendations, a headline such as: The Companies AI Systems Recommend Most Often, is more precise than: The Best Companies in America, unless the research actually supports the broader claim. Accuracy is more important than creating a stronger headline than the data deserves.

Statistics

Statistics should be presented with enough context to understand what they mean. A percentage without a denominator can be misleading. For example: Company A received 80% recommendation coverage, is more useful when readers can also determine:

How many prompts were tested
How many AI systems were included
What market was studied
When the data was collected

The required level of detail depends on the importance of the claim.

Sample Size

Small samples can still produce useful observations. But a small study should not be described as though it represents an entire market with statistical certainty when it does not. Where sample size materially affects interpretation, it should be disclosed.

Methodology Changes

Research methods can improve. If a methodology changes materially, future research may not be perfectly comparable with previous studies. Where appropriate, we should identify:

What changed
Why it changed
Whether historical comparisons remain valid

Methodological evolution is acceptable. Undisclosed methodological drift is less useful.

Original Research

Original research should attempt to create information that did not previously exist in the same form. Examples may include:

AI recommendation datasets
Citation indexes
Industry benchmarks
Longitudinal studies
Original surveys
Competitive datasets

We prefer research that allows readers to understand: what was measured, how it was measured, when it was measured, and what limitations exist.

Attribution

When using another organization's research, data, reporting, or original ideas, attribution should be provided where appropriate. We do not want our publications to create the impression that third-party research originated with 3D Chess Media when it did not. Likewise, we welcome reasonable attribution when other organizations cite original 3D Chess Media research.

Plagiarism

Content should not be copied from another source and represented as original work. Reasonable: Quotation, Citation, Summary, Analysis, and Commentary. may be appropriate when permitted by law and properly attributed. Copying another publication's work without appropriate attribution is inconsistent with our editorial standards.

AI-Assisted Publishing

3D Chess Media may use artificial intelligence as part of research, writing, analysis, editing, classification, data processing, or publishing workflows. AI may assist with:

Drafting
Summarization
Data organization
Classification
Research synthesis
Editing
Formatting
Quality-control processes

Using AI does not remove our responsibility for the information we choose to publish. The appropriate level of human review depends on the subject, risk, and type of content.

High-Risk and Regulated Topics

Some portfolio properties may cover subjects capable of materially affecting:

Finances
Credit
Insurance
Housing
Health
Safety

Those areas may require additional care. Relevant properties may implement stronger processes involving:

Subject-matter review
Source requirements
Disclaimers
Methodology disclosure
Update schedules
Editorial controls

Corporate editorial standards provide a baseline. Individual properties may impose stricter rules.

No Legal Advice

3D Chess Media is not a law firm. Mark B. Huntley, J.D. previously practiced law but does not provide legal advice through 3D Chess Media or its websites. Content discussing: Laws, Regulations, Contracts, and Legal concepts. is informational unless explicitly provided by appropriately engaged legal counsel. Readers requiring legal advice should consult a qualified attorney.

Research First

Our preferred editorial process for data-driven content is:

Define the question
Collect the data
Validate the data
Analyze the result
Identify limitations
Write the conclusion

We do not want to decide the desired conclusion first and then search only for information that supports it.

Conclusions Should Follow the Evidence

Research occasionally produces results that are:

Unexpected
Commercially inconvenient
Contrary to our assumptions
Favorable to competitors
Unfavorable to partners
Difficult to explain

Those results should not automatically be treated as problems to remove. Sometimes the uncomfortable result is the most useful result.

Our General Editorial Principle

The standard we want across The 3D Chess Media Portfolio is straightforward:

Measure what can be measured.
Disclose what should be disclosed.
Separate observation from interpretation.
Preserve disagreement.
Do not claim more than the evidence supports.
Correct material factual errors.
Do not allow commercial relationships to silently rewrite the research.

Questions About Our Editorial Standards

Questions concerning editorial practices, research, corrections, ownership, or disclosures may be directed to:

3D Chess Media, LLC

Email Valerie@3DChessMedia.com

Telephone 949-309-0024

If your question concerns a specific article, study, or portfolio property, please include the relevant URL.

Research & Methodology

Portfolio & Ownership Disclosure