Trust is the absence of doubt: how a website earns it

Dana's kitchen ceiling has a brown stain the size of a dinner plate. It is Sunday evening, and she is on her phone looking for a roofer.

The third site she opens looks fine. Nice photos of slate roofs. A big orange button that says "Get a free quote". She scrolls to the bottom looking for an address and finds a copyright line and an Instagram icon.

Is this a company or one person with a van? Will "free quote" mean someone turning up at her door tomorrow? What happens to her phone number? She does not decide that the site is dishonest. She just cannot answer those questions, so she taps back and tries the next one.

Nothing on that page was false. It simply left her with doubts and gave her no way to settle them.

Dana and her ceiling are imagined. The research, figures and laws below are real.

Trust is missing doubt

We usually talk about trust as something a website adds: badges, testimonials, a confident tone. It helps to flip that around. For a first-time visitor, trust mostly means there are no open questions big enough to stop them.

That framing matters because doubts and reassurances are not symmetrical. One unanswered question ("Is this a real business?") can outweigh five glowing reviews. A visitor does not add up your trust signals and compare the total with a threshold. They look for a reason to hesitate, and if they find one they cannot resolve quickly, they leave.

So the useful design question is not "How do we look trustworthy?" It is "What would make a sensible person hesitate here, and does the page answer it before they have to ask?"

The research below gives that question some structure. Organizational psychologists tell us what people are judging when they decide to trust. E-commerce researchers tell us how that works with a stranger online. Credibility research tells us why only some of your reassurances ever register. And the law, it turns out, already requires many of the answers.

Three questions behind trust

The most cited model of trust comes from a 1995 paper by Roger Mayer, James Davis and David Schoorman. It was written about trust inside organizations, between employees and managers, not about websites. But its structure travels well.

Two details in the paper are easy to miss and useful for anyone designing a page.

First, trust is a willingness to take a risk, not the risk-taking itself. Dana may feel quite positive about a roofer and still not send her details, because the stakes feel high. The answer is not always "be more trustworthy". Sometimes it is "make the step smaller".

Second, the three factors are separate. A site can prove ability brilliantly, with a portfolio of perfect slate roofs, and say nothing about benevolence or integrity. That is what happened to Dana. She did not doubt the photos. She doubted what would happen to her once she handed over her phone number.

Do this: write the three questions at the top of a sheet and list what your page offers for each. Most small-business sites have a long ability column and almost nothing under benevolence and integrity.

Trust between strangers

Mayer's model describes trust that grows over time. A website visitor has no time. They have never met you, there is no history, and the decision often happens in a single visit.

Harrison McKnight, Vivek Choudhury and Charles Kacmar studied exactly that situation. In a 2002 paper on initial trust, they modelled how people decide whether to rely on an unfamiliar web vendor, testing it with a hypothetical legal-advice website. In a companion paper the same year, they built and validated the measurement scales, splitting trust into disposition to trust, institution-based trust, trusting beliefs and trusting intentions.

Three findings are directly useful:

• Perceived site quality mattered. Alongside reputation, the authors describe it as one of the strongest levers a vendor has. A visitor with nothing else to go on reads the site itself as evidence.
• Structural assurance mattered. How safe people felt about the web environment in general fed into their trust in the specific vendor.
• Perceived internet risk lowered intentions to transact. Some of the doubt is not about you at all. It is about the internet, and you inherit it.

That last point is humbling. A visitor arrives already carrying stories about scams and spam. Your page starts slightly in debt.

Do this: list the questions a stranger would ask you in the first two minutes of a phone call. Who are you? Where are you? What does it cost? What happens next? Each one is a section, a line or a link on your page.

Only noticed cues count

Here is the catch. You can answer every question and still lose the visitor, because they never saw the answer.

B.J. Fogg and his colleagues at Stanford's Persuasive Technology Lab spent several years on web credibility around the turn of the millennium. In one large study, published with Consumer WebWatch in 2002, 2,684 people compared two live websites on the same topic and wrote comments about which seemed more credible and why. The most frequently mentioned factor was the "design look" of the site, in 46.1% of comments, ahead of information structure and information focus.

That number is often quoted as "people judge credibility by design". The study measured something narrower: what people commented on when asked to compare two sites. Visual design is the first thing anyone notices, so it gets mentioned. That does not mean it outweighed everything else when people actually decided.

Fogg's explanation for the pattern is a small theory with a big practical use.

The same lab published the Stanford Guidelines for Web Credibility in 2002, based, they say, on three years of research with over 4,500 people. Several read like a checklist for benevolence and integrity: "Show that there's a real organization behind your site", "Make it easy to contact you", "Show that honest and trustworthy people stand behind your site", and "Avoid errors of all types, no matter how small they seem". Nielsen Norman Group's long-running list of four credibility factors, first written in 1999, adds upfront disclosure of things like contact details, fees and return policies.

The guidelines are more than twenty years old. They have aged well, mostly because they describe what people want to know rather than how pages should look.

Do this: for every reassurance on your page, ask "Where will the visitor be when they need this?" Then put it there, not in the footer of a different page.

A map of doubt

Doubts do not arrive all at once. They tend to show up at specific places on a page, usually right before the visitor is asked to do something. Figure 3 is a map of where they typically appear on a service business's landing page, and which of the three trust questions each one belongs to.

A few patterns stand out once you look at a page this way.

Doubt clusters around the ask. Markers 4, 5 and 6 sit within a thumb's scroll of the button. That is where risk becomes real, so that is where the answers belong. "We reply within one working day. No sales calls unless you ask for one" beside a form does more work than a whole About page.

Ability is usually covered, integrity rarely. Most sites show their work (marker 3). Far fewer say what things cost (4) or who is legally behind the site (7).

Some doubts are created by the page itself. A stock photo that looks like stock, a 2021 copyright year, or a broken link are small things, but under prominence-interpretation theory they are noticed and interpreted badly. See Photos that do not look like stock for the photo side of this.

Do this: print your page, give it to someone who has never seen it, and ask them to circle every spot where they would want to ask a question. Then compare their circles with Figure 3.

Doubt at the payment moment

The clearest measurement of doubt in action comes from checkout research, because there the cost of doubt is a lost order.

Baymard Institute regularly surveys US online shoppers about why they abandoned an order. In its 2025 survey of 1,026 US adults, 19% said they had abandoned a checkout in the previous three months because they "didn't trust the site with their credit card information".

Look at the chart again with the trust model in mind. "Didn't trust the site" is the only bar labelled as trust, but several others are integrity doubts in disguise. Extra costs that appear late, a total that cannot be seen up front, an unclear returns policy: each is a moment where the visitor wonders whether the business is playing fair.

Baymard's checkout testing adds a striking detail. Participants often felt that the card fields on a page were less secure than the address fields on the same page, although the whole page was equally encrypted. Their sense of security was a gut reaction to how the page looked. Baymard's recommendation is to visually set the card fields apart, with a border or background used only there, and to place a recognizable seal or icon close by.

Most landing pages for services do not take payment. The lesson still applies to the moment of the enquiry: the phone number field is your card field. It is where a visitor feels most exposed, and it deserves the most explanation. The form tip covers what to ask and when. For which badges people actually recognize, see Trust signals for small businesses.

Do this: find the field on your page that asks for the most personal piece of information. Add one sentence right beside it that says why you need it and what you will not do with it.

What the law already requires

Many of the answers on the doubt map are not optional extras in the EU. They are legal duties. That is good news: the law has already written part of your trust checklist.

This section is general information, not legal advice. Rules differ by country and by what you sell. Check your own jurisdiction or ask a lawyer.

Notice how closely this lines up with doubt 6 on the map. The law and the visitor are asking the same question. A one-line summary beside the form ("We use your details only to reply to this enquiry. Privacy notice.", with the last two words linking to the notice) answers both, as long as the full notice behind the link is complete.

If you sell through a marketplace rather than your own site, the platform now has to check you. Since 17 February 2024, Article 30 of the Digital Services Act requires online platforms that let consumers conclude distance contracts with traders to collect each trader's name, address, phone and email, an identity document, payment account details, trade register number where applicable, and a self-certification of compliance with EU law, to make best efforts to assess whether it is reliable and complete, and to show consumers the name, contact details, register details and self-certification. (Platforms that are themselves micro or small enterprises are exempt under Article 29.) The law's logic is the same as Dana's: before buying from a stranger, you should be able to find out who the stranger is.

That last box deserves emphasis because it is where trust design most often goes wrong. Badges work partly because visitors cannot check them. That is exactly why the law treats a fake one so harshly. If you did not earn the seal, do not show it. The same logic applies to review counts and testimonials, which is the subject of Social proof patterns.

Remove the doubts

Here is Dana's roofer again, as an enquiry card. Each button adds one answer to one doubt from the map. Watch how little extra space it takes.

The business, prices and ratings in the demo are fictional. The point is the proportions: six doubts answered in about seventy words, most of them right beside the button.

Figure 5. The same page before and after answering benevolence and integrity doubts. Fictional example.

What to change tomorrow

1. Run the three-column audit

   Write ability, benevolence and integrity as column headings and list what your page offers under each. Fill the shortest column first. For most small businesses, that is integrity: price, terms and who is legally behind the site.

2. Answer doubts where they appear

   Move "what happens next", "why we ask for this" and a privacy line to within a scroll of your button. Use the doubt map as a checklist and test it with one fresh pair of eyes.

3. Make the step smaller

   Trust is a willingness to take a risk, so reduce the risk. Make the phone number optional, promise no unrequested calls, offer a reply by email. Smaller risk needs less trust.

4. Hunt down doubt you created

   Fix broken links, update the copyright year, replace photos that look like stock, check the page on a phone. Errors are noticed and interpreted badly, which is the worst cell in Figure 2.

5. Cover the legal floor

   Check that your privacy notice has every Article 13 item and is linked at the form, that the site runs on HTTPS, that your identity, address and contact details are easy to find, and that every badge you show is one you are authorized to use.

Dana found a roofer that evening. The site was plainer than the one with the orange button. It said who would come, roughly what it would cost and that they would email first. That was enough. The stain is gone.

Sources

1. Mayer, Davis & Schoorman (1995). An Integrative Model of Organizational Trust. Academy of Management Review, 20(3), 709–734. Consulted for the definition of trust, the three trustworthiness factors and the distinction between trust and risk-taking.
2. McKnight, Choudhury & Kacmar (2002). The impact of initial consumer trust on intentions to transact with a web site: a trust building model. Journal of Strategic Information Systems, 11(3), 297–323. Abstract and summary consulted for site quality, structural assurance and perceived internet risk.
3. McKnight, Choudhury & Kacmar (2002). Developing and Validating Trust Measures for e-Commerce: An Integrative Typology. Information Systems Research, 13(3), 334–359. Abstract consulted for the four trust constructs.
4. Berger & Calabrese (1975). Some Explorations in Initial Interaction and Beyond: Toward a Developmental Theory of Interpersonal Communication. Human Communication Research, 1(2), 99–112. Consulted for the scope and core claim of uncertainty reduction theory.
5. Fogg, Soohoo, Danielson, Marable, Stanford & Tauber (2002). How Do People Evaluate a Web Site's Credibility? Stanford Persuasive Technology Lab and Consumer WebWatch. Summary consulted for the sample of 2,684 and the 46.1% design-look figure.
6. Fogg (2003). Prominence-Interpretation Theory: Explaining How People Assess Credibility Online. CHI '03 Extended Abstracts, 722–723. Abstract consulted.
7. Stanford Persuasive Technology Lab (2002). Stanford Guidelines for Web Credibility. The ten guidelines and stated research base consulted.
8. Harley (2016), Nielsen Norman Group. Trustworthiness in Web Design: 4 Credibility Factors. Consulted for upfront disclosure and the 1999 origin.
9. Holst (2016, updated with 2023 and 2025 data), Baymard Institute. How Users Perceive Security During the Checkout Flow. Consulted for the 19% figure (1,026 respondents, 2025) and perceived-security findings.
10. Baymard Institute (updated 22 September 2025). Cart Abandonment Rate Statistics. Consulted for the full list of abandonment reasons used in Figure 4.
11. Regulation (EU) 2016/679 (GDPR). Official text on EUR-Lex. Articles 12, 13, 32 and 83 consulted.
12. Directive 2011/83/EU on consumer rights. Consolidated text on EUR-Lex. Article 6(1) consulted.
13. Regulation (EU) 2022/2065 (Digital Services Act). Official text on EUR-Lex. Articles 29, 30 and 93 consulted.
14. Directive 2005/29/EC on unfair commercial practices. Consolidated text on EUR-Lex. Annex I points 1–4 and Article 13 consulted.
15. Directive (EU) 2024/825 (Empowering Consumers for the Green Transition). Official text on EUR-Lex. New Annex I point 2a and application date consulted.

Sources checked 18 October 2026. Only publicly available material was consulted; where a paper was paywalled, we relied on the publisher's abstract and say so above. The roofer, the enquiry card and all names in examples are fictional, and the figures are original illustrations.