The Trust lens asks one question: does a visitor who understands your offer have a reason to believe it? PageLint runs eight Trust checks. Some are programmatic (does the page link to a Privacy Policy?), some need a language model (do the testimonials look like your buyers?), and one is a hybrid. None of them measure whether your product is good. They measure whether the page gives a stranger evidence that it is. This article walks through each check, what it flags, and what passing looks like.
Why trust comes after clarity
A visitor has to understand the page before they can doubt it. That is why the Clarity lens runs first. Once the visitor knows what you sell, a second question starts: why should I believe you?
Cialdini's Influence treats this as two separate mechanisms. Social proof says other people like me did this. Authority says someone qualified vouches for this. Both rely on the visitor being able to check, or at least feel they could.
We view a behavior as more correct in a given situation to the degree that we see others performing it.
The Stanford Persuasive Technology Lab studied the other side of this: how people judge whether a website is credible at all. The project's guidelines (B.J. Fogg and colleagues, 2002) include showing that a real organization stands behind the site, making it easy to contact that organization, and making it easy to verify claims. Several of the checks below are that advice turned into something a parser can test.
TR-1: Privacy and Terms links present
The baseline. The page links to a Privacy Policy and Terms of Service, usually from the footer. Both missing is a medium-severity flag, one missing is low.
This is the least glamorous check in the lens and the most common to fail on early-stage pages, because founders ship the hero first and the legal pages never. A visitor deciding whether to enter an email address or a card number looks for exactly these links. Missing ones read as "nobody is accountable here."
The check is programmatic: it looks for links whose href or text point at privacy and terms pages. It does not read the policies. A link to a 404 will pass the parser and fail the visitor, so click your own footer.
TR-2: Testimonials show similarity to the target audience
Featured customers should look like the people you want to convince. Cialdini's similarity amplifier is the basis: we are more persuaded by people who resemble us than by people who impress us.
This is an LLM check. The model is given the page's stated audience and the testimonial block and asked whether the two match. Confidence is high when the audience is clearly stated and the mismatch is obvious, and lower when the page never says who it is for. The longer treatment is in social proof without logos, and the testimonial format itself is covered in testimonials that convert.
TR-3: Social proof present
The simplest social-proof test: does the page show at least one form of it? A testimonial section, a customer-logo strip, or a usage number all count. None detected is flagged.
Pre-launch pages often fail this check. Don't invent proof to fix it. Early products can still show something real: a count of audits or signups that happened, a named beta user with permission, a link to a public build log. A small true number beats a large vague one, and a fabricated one is worse than none. The FTC's rule on consumer reviews and testimonials (16 CFR Part 465, 2024) targets fake reviews directly, so manufactured proof is a legal risk as well as a credibility one.
This check is programmatic and low severity on its own. It matters most in combination: TR-3 passing with TR-2 failing means you have proof, but not proof that persuades your buyer.
TR-4: Authority signals are verifiable
"As featured in" strips, certifications, awards and credentials should link to something a visitor can check. Cialdini's point about authority is that people respond to the symbols of expertise. That is exactly why unverifiable symbols are a problem: the visitor who does check finds a prop.
The check is programmatic. It looks for authority claims and tests whether they link to a source. Claims that cannot be checked are flagged, and a false "as seen on" claim is a deceptive-practice risk under the FTC Act. Cialdini's counterintuitive finding in the same chapter is worth keeping in mind: admitting a small weakness before a strength makes the strength more believable.
TR-5: Familiar trust badges over obscure ones
If the page shows trust badges, they should be ones visitors recognize: PayPal, Visa, Stripe, Google, Cloudflare, or a named standard like SOC 2 or ISO 27001. A generic padlock or a "100% secure" graphic nobody has seen carries little weight and can read as filler.
Baymard Institute's checkout research looks at how users judge payment security and lists distrust of a site among the reasons people abandon a purchase; recognizable brands are the cue users have to go on. The check follows that logic. It has a short list of known badges and flags badge-like images that match none of them.
This is a low-severity programmatic check and only applies when badges exist. Having no badges at all is not a TR-5 failure. If you accept payment through Stripe or Paddle, a small payment-processor mark near the button is a recognizable badge that costs you nothing.
TR-6: Objections handled near the CTA
The moment before a click is when doubt peaks. The check looks for the common objections answered within one viewport of the main CTA: "no credit card required", "cancel anytime", "money-back guarantee", "SOC 2".
Only state what is true. If you have a no-refunds policy, say so rather than implying a guarantee, because a policy found at checkout costs more trust than a visible one costs conversions. The check is LLM-based and major severity: the model reads the text around the CTA and decides which objections it answers.
TR-7: Pricing is transparent
Price visible on the page, or one click away. "Contact sales" with no anchor is flagged for transactional offers. This is a hybrid check: parsing finds a price or a pricing link, and the model judges whether the page is the kind that should show one.
Nielsen Norman Group has written for years about hiding prices on B2B sites, and the finding is consistent: users read missing prices as a sign that the price is high or negotiable, and many leave rather than ask. For a self-serve product the case is stronger still. A visitor who cannot find the price assumes the worst.
If your price varies, show a starting price or a range with what drives it. An anchor is enough to pass.
TR-8: Real human faces present
Team, founder or customer photos should be real, not stock. This is Cialdini's liking principle applied to trust: people believe people they can picture. A named founder photo on the about section does more than an illustrated team of identical smiles.
The check is an LLM evaluation with medium confidence, since judging "stock" from page context is not always certain. It looks at alt text, file names, captions and surrounding copy. Generic stock-photo naming and placeholder captions are the usual flags. The severity is minor because the fix is cheap and the effect is modest. If you are a solo founder, one honest photo and one honest sentence is the whole job.
How the eight fit together
The checks cover three layers. TR-1, TR-4 and TR-5 are about legitimacy: is there a real, accountable organization? TR-2, TR-3 and TR-8 are about people: do real customers and real faces back this up? TR-6 and TR-7 are about the decision itself: are the risks and the price on the table?
Most landing pages fail in the same order. They have testimonials but no similarity (TR-2), badges nobody recognizes (TR-5), and a CTA with no reassurance beside it (TR-6). The fastest wins are usually TR-6 and TR-7, because they are copy edits rather than new assets.
Put the evidence where the visitor's doubt is.
Cialdini's warning applies to the whole lens. Any of these signals can be faked, and visitors, regulators and platforms are all getting better at spotting that. The sleuth approach from Cialdini's seven principles is the working rule: find the real proof you already have, and build the missing proof rather than painting it on.
Check your own page
The free audit runs the Trust checks that matter most against your URL and quotes the exact text it flagged. Run the free audit on your landing page to see which of the eight your page passes.
Sources cited
Primary:
- Cialdini, Robert B. Influence: New and Expanded. Harper Business, 2021. Ch. 4 (Social Proof), Ch. 5 (Authority), Ch. 3 (Liking).
- Fogg, B.J., et al. Stanford Guidelines for Web Credibility. Stanford Persuasive Technology Lab, 2002.
Secondary and corroborating:
- Baymard Institute. Checkout usability research on perceived payment security and trust.
- Nielsen Norman Group. Articles on displaying prices and on credibility in web design.
- Ogilvy, David. Ogilvy on Advertising. Crown, 1983. (Testimonials and the case for stating the price.)
Regulatory:
- Federal Trade Commission. Trade Regulation Rule on the Use of Consumer Reviews and Testimonials, 16 CFR Part 465, 2024.
Further reading:
- Cialdini's 7 principles for landing pages
- Testimonials that convert
- Social proof without logos
- The Clarity lens
- PageLint principles for the full framework reference.