Whether a topic counts as YMYL does not depend on the site's sector, the length of the article or the age of the domain. There is only one criterion: how much harm results if the information is wrong. Google's Search Quality Rater Guidelines define this group as topics that could significantly impact the health, financial stability or safety of people, or the welfare of society, and that is where the abbreviation comes from: Your Money or Your Life.
The guidelines group harm under four headings: health and safety, financial security, society, and other forms of harm that fall outside these. The deciding factor is always the topic itself. A harmful file can be placed on any page, but that does not make the page's topic YMYL; the topic itself has to carry the potential for direct impact.
YMYL is a scale of harm, not a sector label
What is easy to miss: the guidelines explicitly describe YMYL assessment as a spectrum, not an on/off switch. In their own words, most topics are not YMYL, and a topic falls somewhere in the range from clearly YMYL, through maybe, to probably not. The heaviest review burden only applies to pages at the top end of the scale.
In practice, this means two pages on the same site do not have to carry the same burden of proof.
| Page topic | Position on the scale | Rationale |
|---|---|---|
| Heart attack symptoms and when to go to the emergency room | Clearly YMYL | Wrong threshold information can delay seeking care and cause direct physical harm |
| How often to replace a toothbrush | In between | A health topic, but a small deviation causes no measurable harm to anyone |
| How much you need to save for retirement | Clearly YMYL | Bad advice damages a person's ability to support themselves and their family |
| Which channel to use to obtain a tax registration certificate | In between | Procedural information that does not drive financial decisions, but wastes time if it goes out of date |
| When a vitamin was discovered | Generally not YMYL | Information that does not lead to decisions |
A health site does not need a reviewing expert on every page. Concentrating review resources on pages at the top end of the scale is both more defensible and cheaper than handing out a superficial badge across the entire archive.
Where Google actually wants reviewer information
Among the self-assessment questions in Google's helpful content documentation is this one: was this content written or reviewed by an expert or enthusiast who demonstrably knows the topic well? The industry usually overlooks two details in that sentence. First, writing and reviewing are presented here as alternatives, not as two overlapping requirements. Second, alongside the expert, an "enthusiast who demonstrably knows the topic" is explicitly accepted too.
The same documentation gives clear advice on authorship: it recommends adding accurate author information and bylines where readers would expect them, with the byline leading to a page that gives details about the author. There is no corresponding recommendation for the reviewer.
In the full text of the Quality Rater Guidelines, the phrases "reviewed by" and "reviewer" do not appear at all. The guidelines ask raters for something quite different: to find out who is responsible for the site and who created the content on the page. If this information is insufficient on a topic requiring a high level of trust, the page gets a low quality rating. In other words, what is mandatory is not a reviewer label but being able to tell who is responsible. The site's About page carries much of that burden.
What the review badge is really worth
The "Reviewed by: Dr. X" line at the top of a page is not a ranking signal documented by Google. Two separate statements support this. Google states that E-E-A-T itself is not a specific ranking factor. On quality raters, it says that raters have no control over how pages rank and that rater data is not used directly in ranking algorithms. The guidelines are not a scoring formula but a reference text describing what the systems are looking for.
This does not mean the badge is useless. Its function simply lies elsewhere. Readers see who stands behind the page. A third-party verifier can test the claim. Inside the organization, one person becomes genuinely accountable for the accuracy of the text. None of these is an algorithmic shortcut; all of them produce verifiability. Content trustworthiness is also not the only lever of search engine optimization, just one signal among many.
A common flawed line of reasoning goes like this: major health and finance publications show the reviewer at the top of articles, those publications rank well, so the badge brings rankings. Those same publications also differ in institutional resources, editorial staff, brand recognition and link profile. There is no public study measuring the badge's contribution in isolation from these variables, so reading the relationship as causation has no basis.
The difference between a cosmetic badge and a real review
There is no single way to show a review process, and the different ways are not equivalent. The comparison below separates what each option actually proves to a reader and to an independent verifier.
| Display format | What it actually proves | Can it be tested externally? | Real weight |
|---|---|---|---|
| Anonymous "expert approved" label or badge image | Only that a claim exists | No | None |
| Reviewer's name and title | That a name has been given | Partly, the name can be searched | Low |
| Name, bio page and verifiable credentials (registration, license, institution, publications) | That the person is competent in that field | Yes, through third-party records | Medium |
| Review note: which section was approved, what was corrected | That the review actually took place | Yes, through the text itself | Medium |
| Published editorial policy, source list and corrections log | That the process depends on the organization, not on one person | Yes, repeatable from page to page | High |
The distance between the first and last rows shows where most "we implemented reviewed by" efforts actually stand. The first row is a design element and proves nothing. From the third row onward, a testable claim emerges. The reviewer's bio page becomes decisive here and is held to a higher evidentiary bar than an author bio; credentials such as a license, registration or institutional record need to be verifiable outside the site.
Which node do reviewedBy and lastReviewed belong to?
A common technical mistake is writing the reviewedBy property on the Article node. In the schema.org vocabulary, the domain of the reviewedBy and lastReviewed properties is WebPage. author, by contrast, is defined on CreativeWork; Article is a CreativeWork but not a WebPage. So author belongs on Article, while reviewedBy does not.
The correct setup is to mark up the page as WebPage or one of its subtypes, attach the review to that node and link the article as the page's main entity. For health content, MedicalWebPage can be used, because that type is a subclass of WebPage and inherits both properties.
{
"@context": "https://schema.org",
"@type": "MedicalWebPage",
"url": "https://example.com/health/home-blood-pressure-measurement",
"lastReviewed": "2026-05-02",
"reviewedBy": {
"@type": "Person",
"name": "Dr. Firstname Lastname",
"jobTitle": "Cardiologist",
"url": "https://example.com/team/firstname-lastname",
"sameAs": "https://medical-association-example.org/member/12345"
},
"mainEntity": {
"@type": "Article",
"headline": "How to measure blood pressure at home",
"author": {
"@type": "Person",
"name": "Author Firstname Lastname",
"url": "https://example.com/team/author-firstname-lastname"
}
}
}Two small details matter. lastReviewed takes a date, while dateModified takes a timestamp that can include time; they are not the same field and are not interchangeable. The review date should also only be moved forward when a genuinely new review has been carried out, because refreshing the date without touching the text amounts to declaring work that was never done.
A word of caution on expectations: reviewedBy does not appear anywhere in Google's Article structured data documentation. This markup does not produce a rich result. The gain is that the visible statement on the page and the machine-readable statement do not contradict each other. Structured data is a consistency tool, not a promise of rankings.
How to set up the process for health and finance content
A concrete example helps. A page explaining how to measure blood pressure at home sits at the top end of the scale, because a wrong threshold value could lead a reader to put off seeing a doctor. The meaningful workflow here is: the content team writes the text, a physician in the relevant field checks the values and warning thresholds, the corrections are worked into the text, and the physician's name on the page links to a verifiable bio. The reviewer's area of expertise must match the page's topic; a dermatologist's signature under a cardiology article does not deliver the assurance the reader is promised.
On the finance side, the distinction works a little differently. The question "how much should I save for retirement" sits at the top end of the scale and deserves review by a certified accountant or a licensed professional. "Where do I get a tax registration certificate", on the other hand, is procedural information; what needs checking here is not expert opinion but whether the procedure is current. Two pages on the same site do not have to go through the same process.
The places a reviewer actually needs to look are not the same on every page either. Four items work well in practice:
- Numerical thresholds, dosage-like critical values and ratio calculations
- Statements that contradict or go beyond established expert consensus
- Points where the reader should be directed to professional help but is not
- Sections that may go out of date because they depend on regulations or current practice
What can sites without in-house experts do?
Most small teams do not have a physician or certified accountant on staff. That does not mean nothing can be done in YMYL areas, but it does mean the easy route is closed. A fabricated reviewer profile is the most harmful option, because the guidelines put fake author and reviewer profiles and false credential claims straight into the lowest quality rating. Exaggerated or mildly misleading claims of expertise also earn a low rating, and the guidelines give raters a clear instruction here: the assessment should be based not on the site's own claims of expertise but on the content itself, reputation research and verifiable credentials.
The remaining options are boring but effective.
- Narrow the scope. Not answering questions at the top end of the scale at all and directing readers to the primary source is a legitimate editorial decision. What a page chooses not to say is also a choice.
- Make the basis visible. Linking claims to primary sources such as official bodies, professional associations or peer-reviewed publications makes the basis testable without claiming your own expertise.
- Explain the method. When you write down how the information was compiled, what was tested and what was left out of scope, readers can evaluate the process. Google's "how" question asks for exactly this.
- Use first-hand experience in the right place. The guidelines say that even on clearly YMYL topics, pages sharing lived experience can be considered high E-E-A-T as long as the content is trustworthy and consistent with established expert consensus. Experience is no substitute for recommending treatment options, but it is a legitimate basis for describing how to cope with the process.
- Get a one-off external review. Having an independent expert review a few pages at the top end of the scale once, rather than every article, is both more honest and easier on the budget than handing out names across the whole archive.
Three questions that often come up in practice
Can the author and the reviewer be the same person?
Yes. Google's self-assessment question asks whether the content was written or reviewed by an expert; it does not require both. Adding a separate reviewer line to text written by someone competent in the subject only adds an empty layer. A separate reviewer makes sense when the author is not a subject-matter expert.
How often should the review date be updated?
There is no documented interval. The criterion is not the calendar but how volatile the topic is. Sections that depend on regulations, prices or clinical practice are revisited when those change; permanent definitions do not need frequent review. Moving the date forward without touching the text, on the other hand, is indefensible, because it declares a review that never happened.
Does adding reviewedBy change anything in search results?
Not in terms of a visible rich result. Google's Article structured data documentation does not list this property, so there is no search feature tied to it. What the markup gives you is that the page's visible statement and its machine-readable statement say the same thing. That consistency is easy to establish if a real process already exists; if it does not, the markup will not create one.
