Free tool · Conversion pages
Free advertorial claim risk scanner
A claim is risky when it asserts a fact and points at nothing. This reads a landing page sentence by sentence, separates the claims from the copy, and reports which ones cite a source a reader could open, flagging health, financial and comparative claims separately because they are held to a higher standard. It checks sourcing, not truth, and it is not legal advice.
The claim list is free and unlimited, with no account, because it runs on rules rather than a model. Reading the sources you cite needs a free sign-in and is limited to one run a month, because it costs us money.
AEO depth layer
See the Source check
The Source check is computed from your own page and costs nothing to run. We ask for a free account because the expensive tools on this shelf stay free, and because it lets us tell you when your result changes. No card, no trial clock.
- We open the pages your claims link to and read them.
- Each cited claim comes back with whether that source actually supports the sentence, or only looks like it does.
- It checks sourcing, never truth, and a source that agrees with you can still be wrong. This is not legal advice.
Signing in saves the details you entered above so we can follow up about your results. See our privacy policy for what we keep and why. The Advertorial Claim Risk Scanner result you already have stays free either way. Privacy policy.
What each verdict means, and what it does not
Four outcomes, and the distinction between the middle two is the one people get wrong. A claim we could not check is not a claim that passed.
| Verdict | What it means | What to do |
|---|---|---|
| No source | The sentence asserts a fact and the paragraph links nowhere a reader could check it. | Link the source, rewrite it as your position, or cut it. |
| Unconfirmed | It cites something, but the source was not read. Abstention, not approval. | Sign in and run the source check, or open the link yourself. |
| Higher standard | A health, financial or comparative claim. Escalated whatever its source says. | Make sure you can defend it, and carry the disclaimer your category needs. |
| Source agrees | The page it links to supports the sentence. Not proof the sentence is true. | Check the source is one you would be happy to be judged on. |
The three ways a page looks sourced and is not
Each of these passes a glance and fails a reader who follows the link. The scanner catches the first two and is honest that it cannot catch the third.
- The citation that points at the same claim. A sentence links to an article that asserts the same thing with no source of its own. Structurally it is cited; in substance nothing has been shown. The source check is what finds this, because it reads the page at the other end.
- The link that is an ad. "Learn more" pointing at a partner reads as a citation to a scanner that only counts links. This one does not count it, on the anchor text, which is also why a link labelled "Source" does count.
- The footnote with nothing linking to it. A superscript number and a reference list at the bottom is a real citation and this tool misses it, because nothing connects the sentence to the note. Pages doing that will look less sourced here than they are, and that is a limitation rather than a verdict.
The other direction of the same problem is a page built from claims nobody checked in the first place. Presell generates the page from evidence rather than generating the page and looking for support afterwards, which is why its claims carry a source and a date before anybody scans anything.
How to check the claims on a landing page
Step 1
Enter the page your ads point at
An advertorial, listicle, presell page or product landing page. The claims that matter are the ones a buyer reads immediately before deciding, which is exactly this page rather than your homepage.
Step 2
Read the unsourced list first
Every sentence that asserts something and links to nothing. This is the list a reader, an ad reviewer or a regulator would build by hand, and it is free and unlimited here.
Step 3
Look at what got flagged as a higher standard
Health, financial and comparative claims carry more risk than the rest of the page. They are not necessarily wrong, they need to be defensible, and this tells you which sentences those are.
Step 4
Then check the sources you did cite
Sign in and the deeper layer opens the pages your claims link to and reports whether each one actually supports the sentence. A citation pointing at another unsourced page is the most common way a page looks sourced and is not.
When you outgrow this tool
Finding an unsourced claim is the easy half. Presell builds the page from evidence in the first place, so every factual sentence carries the source and the date it came from.
Finding an unsourced claim is the easy half. Presell builds the page from evidence in the first place, so every factual sentence carries the source and the date it came from.
FAQ
Advertorial Claim Risk Scanner questions
What is checkable about a claim from the page alone, what is not, and why an unsourced claim is reported as unsourced rather than as false.
Does this tell me whether my claims are true?
No, and nothing can do that from the outside. It answers a narrower question that is genuinely checkable: does this sentence assert a fact, and does it point at a source a reader could open and read. A true claim with no source still comes back as unsourced, because from the page alone there is no way to tell the difference between a true claim and a confident one. That is also how an ad reviewer reads it.
Is this legal or compliance advice?
No. It is a structural check on sourcing, and it does not know your jurisdiction, your industry's rules, your substantiation file or what you can produce if asked. Treat the output as a list of sentences worth a second look, not as a clearance. We say the same thing inside the paid product: every factual claim on a Presell page is source-linked, and here is the source. We never claim a page is compliant.
Why is a claim that links to a source still flagged sometimes?
Three reasons. A health, financial or comparative claim is escalated whatever its source says, because those are held to a higher standard. A claim whose source we did not read comes back unconfirmed rather than approved, because abstention is not approval. And a claim whose linked page does not actually support it comes back unsupported, which is the finding the source check exists to produce.
How does it decide something is cited?
The paragraph has to link to a different domain, with anchor text that reads like a citation rather than like navigation. 'The 2026 trial' counts. 'Learn more' and 'Shop now' do not, because a link to a partner is an ad rather than a source, and counting it would make the average sales page look well sourced. This is a heuristic about citation habits, not proof: it cannot tell a citation from an affiliate link, and it misses footnote-style citations where nothing links the sentence to the note at the bottom.
Why does the same page always give the same answer?
Because the classifier behind it is pinned to deterministic decoding and carries a version. That matters more in public than it does inside an account: a customer re-checking a page has a history and can see what changed, while a stranger who runs a free tool twice and gets two answers has only the second one. When the grading behaviour changes, the version changes with it, so old verdicts are never quietly compared against new ones.
What does the free part cost me, and what needs an account?
The free part is the whole claim list, unlimited, no account. It runs on rules rather than a model, so it costs us nothing and there is no reason to meter it. What needs a free account is the source check, because opening and reading every page you cite does cost us money. That is one run per account per calendar month.
What do I do about the claims it flags?
Three options per sentence, and they are all legitimate. Link the source you already have. Rewrite it as your own position rather than as a finding, which is a copy change rather than a compliance one. Or cut it, which is the right answer more often than people expect, because an unsupported claim usually is not what is doing the converting anyway.
More free tools
Every tool on the shelf is free, with no card and no trial. Most need no account at all. See all free tools.
Build a complete competitive battlecard your reps will actually use, and export it as markdown.
Model the pipeline a better win rate is worth, and what manual competitor research is already costing you.
Assemble a free competitor monitoring stack: prefilled alerts, feed URLs, and a review cadence you can set up in ten minutes.
Build the JSON-LD properties that make a page attributable rather than eligible for a rich result, with the reason each one matters.