Someone has quoted you for review markup, or an audit flagged that you have none. Before you buy it, work out which of your pages could carry it at all, because for most advisory and accounting firms the honest answer is one page or none.
The confusion is fair. Review markup is real, it is well documented, and it does produce the gold stars everybody wants. It simply does not do that on a page where the firm being reviewed is the firm publishing the page.
What follows is the eligibility question first, then what the markup needs when it is allowed, then where the stars you see in search really come from. The tool half comes last, and it holds a surprise.
What Review Markup Was Supposed to Do
Structured data is a set of labels you add to a page so a search engine can read it without guessing. Review markup labels two things, and the gap between them is the part everyone skips. Which of the two does the code on your site claim right now?
- A review: One person's view of one thing, with an author and a rating.
- An aggregate rating: The summary of many reviews, carrying an average and a count.
Both exist so a search result can show a star row and a review count under the title. That row lifts click-through, which is why the markup gets sold hard, and why so many firms have been quoted for something they cannot use.
The rules were tightened years ago and have not loosened since. Most of the advice still online was written before that change.
Why Your Own Pages Cannot Show Stars
Google's position is that when the thing being reviewed controls the reviews about itself, pages using the local business or organisation types are not eligible for the star feature. That covers your services page, your about page and any testimonials page you publish.
We have set this out at length elsewhere, so it is worth two sentences here and a pointer. The reasoning and the exact wording sit in our piece on local business schema, and what to do with the quotes you have already collected is covered on the client testimonial page.
The point worth carrying forward is narrower and more useful.
- What triggers the exclusion: Who controls the reviews being described.
- What has nothing to do with it: The quality of your code, the plugin you used, or whether the reviews are real.
Change what the page is about and the rule changes with it. That is the whole of the next section.
Which Page Types Are Actually Eligible?
Google publishes the list, and it is longer than most firms expect.
Seventeen types can carry it. Most of them describe media, retail or events, so the list runs through books, courses, films, games, recipes, music, software and dated events.
Local business and organisation sit on that list too, and are still excluded when the reviews are self-controlled. That pairing is why reading the list alone misleads people.
Run your own site against it and the field narrows fast.
| Type to use | What the page has to actually be | Does a firm plausibly own one? |
|---|---|---|
| Course | A course or seminar you deliver, reviewed by attendees | Sometimes, if you run real training |
| Event | A dated event, reviewed afterwards | Sometimes, for a workshop or a seminar series |
| Product | Something sold with a fixed scope | Rarely, and a retainer is not a product |
| Book | A book you published | Occasionally, for a partner who has written one |
| Software application | A tool or calculator you built and give out | Rarely |
| Local business or organisation | Your firm | Yes, and excluded when you control the reviews |
A paid course with real attendee reviews is the most common lawful route for a professional firm. A fixed-scope engagement dressed up as a product is the most common unlawful one, and the two look alike on a quote.
One oddity worth knowing, because it will confuse you: The how-to type sits on the eligible list while the how-to search feature itself was withdrawn, so markup can be valid and still produce nothing at all. Valid and useful are separate questions, and only the first one is what a validator answers.
What the Markup Has to Contain
Assume you have a page that qualifies. Two rules catch people out, and both sit in Google's own guidance. Which of your pages would pass them today?
| What it is | Property to add | Requirement |
|---|---|---|
| The average score | ratingValue | Always required |
| How many ratings | ratingCount | At least one of these two |
| How many written reviews | reviewCount | At least one of these two |
| The reviewer | author | Required on an individual review |
Google's review snippet documentation states that "at least one of ratingCount or reviewCount is required", alongside the rating value. A block carrying an average with no count behind it fails.
The second rule quietly kills more of these than any coding error does. Google says to "make sure the review content you mark up are readily available to users from the marked-up page", and that "it must be immediately obvious to users that the page has review content".
Read that as a design instruction. Markup describing reviews that a visitor cannot see on the page is the problem, whether the reviews are real or otherwise.
The version that fails silently
A page carries a rating in its code and shows the reviews in a tab, a slider or a pop-up that loads on click. The markup validates, the reviews are real, and nothing appears in search.
- Safe: Reviews printed on the page, visible on load, no interaction needed.
- At risk: Anything behind a tab, a carousel, an accordion or a click.
Where Do the Stars in Search Actually Come From?
Look closely at the results that show stars for a service business and one of three things is happening.
- A review platform is ranking, and not the firm: The stars belong to the platform's own page about the firm, on the platform's domain, where the platform does not control the reviews it hosts. That is a legitimate use, and the page earning the click is theirs.
- The listing is a Google Business Profile, and not a web page: Ratings in the map pack and the knowledge panel come from Google's own review system and involve no markup by you at all. Earning those is a different job, covered in how to ask clients for reviews and how to respond to reviews once they arrive.
- The page really is about a product or a course: Which brings you back to the eligibility table.
Does that mean the platforms are your competitors in search? For your firm's name, frequently yes, and there is not much to do about it beyond being the better answer for everything else. Where a profile carries something untrue, a fake Google review has its own removal route.
Chasing star markup for a services page is effort spent on the one result you cannot have, while the ratings a prospect will actually see sit in the map pack and are earned by asking.
The Testing Tool Most Guides Still Name Is Gone
Search for how to check this and half the results will send you to the Structured Data Testing Tool. It no longer does what those guides describe.
Google removed the Google-specific validation from that tool and moved it to a new domain, where it became the Schema Markup Validator. Two tools now cover what one used to, and they answer different questions.
| Tool | What it answers | What it will not tell you |
|---|---|---|
| Rich Results Test | Which Google rich results your markup can generate | Whether the schema is valid in general terms |
| Schema Markup Validator | Whether the markup is valid schema.org syntax | Whether Google will ever show anything for it |
| Search Console reports | What Google actually found across the site over time | Anything about a page you have not published yet |
Google's own structured data documentation describes the Rich Results Test as "the official Google tool for testing your structured data to see which Google rich results can be generated by the structured data on your page", and the validator as one that will "validate all Schema.org-based structured data that's embedded in web pages, without Google feature specific warnings". The validator's move to schema.org was announced by schema.org itself.
Which one should you use? For a review question, the Rich Results Test, every time. A green result in the validator tells you the syntax is well formed and tells you nothing about whether a star will ever appear, and that gap is exactly where a firm concludes the markup worked.
What a Firm Should Do About Reviews Onsite
Nothing big, and the useful version takes an afternoon.
- Leave existing review code alone: Unused markup does no harm in search. It simply produces nothing, so there is no cleanup emergency.
- Stop paying for star markup on service pages: The feature is closed to those pages, so the work cannot do what it promises.
- Put the markup where it qualifies: If you run a paid course or a ticketed seminar with real attendee feedback, that page is your one genuine candidate.
- Show the reviews you quote: Wherever proof appears, make it visible on the page without a click, which serves readers whether or not any markup is involved.
- Test with the right tool: Run the Rich Results Test on the page in question before and after any change.
The broader on-page work that does move results is set out in our on-page SEO checklist, and review markup sits a long way down that list for a firm of this shape.
Where This Leaves Your Schema
Everything here reduces to what the page is actually about.
If the answer is your firm, review markup will not produce stars there, and no amount of correct code changes that. If the answer is a course, an event, a book or a genuine product, you have a candidate, and then the requirements are a rating value, a count and reviews a visitor can see without clicking anything.
- Test it with the Rich Results Test, because the validator answers a different question.
- Treat a valid result and a working result as separate findings.
If you would rather have someone tell you which of your pages qualifies and what your competitors are actually doing to get those stars, our SEO services start with a review of what your site already publishes.
Frequently Asked Questions
Can we mark up feedback we collected by email?
How you gathered it is not the issue. The problem is that your firm controls those quotes, so the pages carrying them stay outside the star feature no matter how the feedback arrived or how carefully it is labelled.
Does a high average rating help us at all if it never renders?
It helps the humans reading the page, which is a real return. What it will not do is change how your result looks in search, so budget for it as trust content and never as a search feature.
How often does this need re-checking?
Re-run the Rich Results Test after any template change, any plugin update and any redesign. Those are the three events that quietly strip markup out of a page without anyone noticing.
Do the third-party platforms charge for this?
Many do, and the star row you see belongs to their page rather than yours. Weigh that carefully, since you may be paying to strengthen a result that outranks you for your own name.
What about ratings shown inside a client portal?
A portal sits behind a login, so search engines never reach it. Nothing there is eligible for any search feature, and nothing there needs markup of any kind.