Make claims and authorship easy to check

Classify the claim

Not every claim needs the same proof. A sentence about what happened on a project, a measured result, a statement of a platform’s behavior and a recommendation are four different things, and a reader should be able to tell which they are reading. Google’s people-first guidance asks whether content shows clear sourcing, evidence of the expertise involved and background about the author or site, and whether it contains easily verified factual errors.^[1] Classifying claims is the practical way to meet that standard.

Mark opinions as opinions. “We recommend a static build for sites under ten pages” is a judgment, and it is stronger for saying what it rests on: the number of projects, the maintenance experience, the cost difference. A recommendation labeled as a recommendation is more credible than one disguised as a fact.

Check the source behind the source

Statistics travel without their conditions. A figure appears in a vendor blog, is repeated in a roundup and becomes “research shows” within a year. Before you repeat a number, find the original and read what was actually measured: the sample, the dates, the method and the limits.

The same discipline applies to platform behavior. Prefer the platform’s own documentation, check its date, and when two sources disagree, say so. If you cannot find the original for a claim, either drop it or label it as unverified. Google also reminds creators that attribution is not a substitute for effort: crediting other sources does not replace the need for original work.^[1] A page that is a tidy summary of three other pages, however well cited, has still added little.

Describe what you actually did

Google’s guidance asks about the “How” behind content. Its example is product reviews: it builds trust when readers understand the number of products tested, the test results and how the tests were conducted, accompanied by evidence of the work such as photographs.^[1] Its separate guidance on reviews adds first-hand evidence in the form of visuals, audio or links, and quantitative measurements.^[3] The same applies to any page that reports results. State what you did, in enough detail that a reader could judge it.

The limits line matters more than it seems. A claim with stated limits is easier to trust than one without, because the reader can see that you looked for the edges. It also protects you: a fair summary of a small test cannot be quoted as a universal law without distorting what you wrote.

Represent the author honestly

Google strongly encourages accurate authorship information, such as bylines where readers would expect them, and bylines that lead to further background on the author and the areas they write about. It warns against deceptive authorship: fabricated creator profiles, such as AI-generated headshots, made-up names or false credentials, are a form of deception that makes a page untrustworthy to both people and its automated quality systems.^[1] It also advises disclosing automation or AI involvement where readers would reasonably expect it, with details of how it was used and why.^[1]

The rater guidelines describe three places to look when judging whether a site or creator is trustworthy: what the site or creator says about themselves, such as an About page; what others say, such as independent reviews, references and news; and what is visible on the page itself.^[4] They also say trust is the most important member of the E-E-A-T family, because an untrustworthy page has low E-E-A-T however experienced or expert it seems.^[4] Google adds that E-E-A-T is not itself a ranking factor, while its systems give more weight to strong E-E-A-T for topics that could significantly affect health, financial stability or safety.^[1]

Nielsen Norman Group’s early web research made a related point: credibility is important because it is unclear who is behind information on the web, and it can be increased by high-quality graphics, good writing and outbound links that show the authors have done their homework.^[5] An author page that says who wrote the advice, what they have done and how to reach them is the minimum. For topics that could affect someone’s health, money or safety, add a qualified reviewer and say what they reviewed.

Maintain a correction path

Pages age and people make mistakes, so a trustworthy site shows how it handles both. Put a “last reviewed” date on pages with time-sensitive claims and update it only when someone has actually reviewed the content. When you change a fact, add a one-line note saying what changed. Give readers an easy way to report an error, such as an email address on the page, and a rule for answering: acknowledge, check, correct, record.

The log is also your protection if an AI answer or another site repeats something you have since corrected. You can show what the page said, when it changed and why. Entries here are illustrative; keep your own.

Apply it to your website

• Each important claim is classified as measured, observed, sourced or opinion, and presented as such.
• Numbers are traced to the original source, with its sample, date, method and limits read and stated.
• Pages that report results include a method note and the material a reader can inspect.
• Authors are real, named and linked to a page describing their experience. No profile or credential is invented.
• AI or automation involvement is disclosed where readers would expect it, with a short explanation.
• Time-sensitive pages carry a last-reviewed date, and corrections are logged and visible.
• Pages are original work and not only tidy summaries of other pages.