Build confidence without pressuring customers

In Chapter 7, you looked at the questions someone may ask before deciding whether to buy. A clear answer to a question can help. Yet a claim alone does not give a person a good reason to believe it. If a page says “easy,” “trusted” or “best,” the reader may still need to know easy for whom, trusted on what basis and best by which measure.

This chapter shows how to support claims with suitable evidence, explain important limits, use reviews responsibly and design a choice that a person can understand and decline. You will build a blank claim record, inspect a page for avoidable pressure and practise writing helpful explanations. No invented customer, result or testimonial is needed. The diagrams make the relationships visible, and each one is explained in words. Two optional videos appear near the end; the written lesson is complete without them.

By the end, you should be able to: tell an objective claim from an opinion; ask what proof a claim needs; recognise why a review cannot prove every kind of result; check whether a price and choice are clear; and revise a message so that it helps a reader decide for themselves.
A08-01-where-this-fits
Figure 1
Figure 1 in words: The previous chapter identified the person's question. This chapter asks four questions about your answer: What exactly are you claiming? What supports it? What limits matter? Can the reader accept, reject or delay the next step? These are checks on your communication, not a way to label a person as trusting or distrustful.

What does confidence mean here?

Confidence in a decision means having enough understandable, relevant information to judge an option and its consequences. It does not mean certainty that nothing can go wrong. A careful person may be confident enough to decide “not suitable for me.” A business can support that outcome by explaining fit and limits honestly.

Trust and confidence are related but not identical. Trust is a person's willingness to rely on someone or something despite some uncertainty. You cannot simply place a “Trust us” badge on a page and assume that trust exists. A reader may look for a working service, clear terms, current content, an accountable organisation and evidence from outside the seller's own statements. Nielsen Norman Group's web usability work describes four factors in perceived website trustworthiness: design quality, up-front disclosure, comprehensive and current content, and connection to the wider web (Harley, 2016). This is a helpful website review lens, not a guarantee that a particular visitor will buy.

Persuasion means trying to influence a decision through reasons, presentation or social cues. It is part of marketing. An informative comparison can persuade because it makes a relevant difference clear. Pressure makes a choice harder to consider freely: hiding a fee until the last step, pretending that time is running out, making decline hard to find or exaggerating what others have done. The distinction is not whether a page has a call to action. It is whether the person can understand the material facts and choose without being misled or cornered. The Competition and Markets Authority (CMA, 2022) describes how online choice architecture, the way digital options are presented, can help or harm decisions.

The practical goal is informed choice. Show the result the offer is intended to deliver, the evidence that supports your statements, the conditions under which they hold, the total cost and a fair next step. A readable page can still be persuasive. It earns that persuasion by helping the person judge the offer.
A08-02-four-supports
Figure 2
Figure 2 in words: A clear answer addresses the real question. Suitable evidence supports the exact claim. Visible limits show when the claim may not apply. Genuine choice gives a person an understandable path to proceed, wait or decline. A gap in one support can weaken the others; an attractive layout cannot replace missing proof.

Start with the reader's actual question

Chapter 7 asked you to name one decision and its information needs. Keep that map nearby. If a reader asks “Will this work with what I already use?”, a page full of broad praise does not answer. If the question is “What will it cost altogether?”, a monthly headline with unavoidable fees disclosed later is incomplete. A person considering a consequential choice may need more detail than someone making a familiar, reversible one.

Before drafting a message, write the question in plain language. Then write the smallest answer that would be truthful and useful. Add the explanation and source that a reader could reasonably inspect. Include a limitation close enough to the claim that it changes its meaning. You are not required to promise that every reader will have the same result.

This sequence also tells you when to stop. If the evidence is missing, do not make a stronger claim while you wait for research. You can describe an intended use or a verified feature more narrowly. A claim is what your communication asserts or leads a reasonable reader to infer. It includes the headline, image, caption, badge, comparison and what the whole page implies, not only the literal words in one sentence. The UK Committee of Advertising Practice (CAP, 2026b) advises that objective claims should be supported by evidence for the meaning consumers are likely to take from them, rather than only the meaning the marketer intended.

What kind of claim are you making?

An objective claim is capable of being checked against evidence: a price, feature, availability, measurable outcome, ranking or comparison. The CAP Code requires documentary evidence before publication for objective claims that consumers are likely to regard as capable of substantiation (CAP, n.d.). Evidence should fit the claim. A feature list can establish that a feature exists; it does not by itself establish that the feature saves every user time. An internal test can say something about the tested conditions; it cannot automatically support a universal promise.

A subjective statement expresses opinion or preference. “I like this design” is different from “This is the fastest design.” But a phrase that sounds like opinion may still imply a factual superiority in context. The label you give a claim does not control how a reader understands it. If a headline, chart or visual makes a concrete assertion, treat it as an assertion and check it.

A comparison needs a defined basis: compared with what, on which measure, for which period and under which conditions? “Better” has no useful meaning until these questions are answered. A superlative such as “number one” can imply a measurable rank. A prediction says something about a future result and needs an honest account of uncertainty. A guarantee commits the business to a result or remedy; it needs precise terms, not a decorative label.
A08-03-claim-path
Figure 3
Figure 3 in words: Draft the statement, ask what a reader is likely to understand, identify its evidence, place necessary qualifications beside it and check the final page. If the proof does not fit, narrow the claim or hold it back. The review includes images and surrounding text, because those can change the apparent meaning.

The same checking habit applies to an attractive number. Ask what was counted, when, across whom and with what denominator. A percentage without a baseline may create a larger impression than the data warrants. A result from a small voluntary survey may tell you what respondents reported, not what all customers experience. If you have no sound number, an accurate plain-language description is better than a fabricated statistic.
A08-04-claim-types-table
Figure 4
Figure 4 in words: A feature claim needs a current specification or demonstration. A measurable benefit needs a method and relevant results. A comparison needs a defined competitor set and measure. A testimonial needs genuine experience and permission, while any factual claim inside it still needs appropriate proof. The image lists questions to ask; it does not claim that any specific offer has passed them.

Evidence has to match the claim

Think of substantiation as the supporting record you would be able to show if the claim were questioned. It can include a tested specification, a documented process, a valid study, current terms or another suitable source. The required strength depends on what is being promised and how consequential a wrong claim would be. CAP's guidance says a customer testimonial alone will not usually substantiate an objective efficacy claim (CAP, 2026b).

Use a simple four-part record:

  1. Exact words and placement. Save the headline, surrounding text and image as the reader will see them.
  2. Likely interpretation. Ask a person unfamiliar with the offer what they think it means. Record their answer without coaching them.
  3. Supporting record. Name the source, owner, method, date and the specific part that supports the statement.
  4. Conditions and review date. Note exclusions, scope, expiry and who will check whether the claim remains true.
Do not treat a citation as a magic stamp. A published research paper can support a general explanation of a framework; it cannot prove the performance of your own service. A customer's positive account can describe their own experience; it cannot establish a typical outcome without a sound basis. A screenshot of a dashboard may show a number at one point in time; it may hide the definition of that number. The more specific the claim, the more specific the proof must be.

A qualification is information that limits or explains a claim. It may say who qualifies, what is included, what is excluded or how long a result took. Place material qualifications where readers will see them with the claim. A tiny note at the foot of a long page is not a fair correction to a bold headline. CAP's Code says qualifications must not contradict a misleading main claim or hide significant limitations (CAP, n.d.).

If your records change, the claim can become stale. Assign a review owner to prices, product features, availability, rankings and any timed offer. Record the last check date. An evergreen explanation of a concept may last longer, but a live offer needs a more frequent check. Never silently turn “planned” into “available now.”

Make the website itself understandable

Trustworthiness is not only about prose. A site can undermine an accurate message if the navigation is confusing, the contact route is hidden or a page contradicts the checkout. Harley's four factors are a useful audit: design quality, up-front disclosure, comprehensive and current content, and connection to the rest of the web (Harley, 2016).

Design quality means information is organised so a reader can find and use it. It is not simply decorative polish. Clear headings, functional links, readable type and an accessible form help a person assess the offer. Up-front disclosure means important terms, identity, costs and next steps are not held back until commitment. Comprehensive and current content means the relevant questions are answered and old details are corrected. Connection to the wider web may include verifiable references and an identifiable organisation, not fabricated badges or meaningless logos.

Ask a reader to locate a specific answer without leading them. Observe whether the page supplies it, whether two pages disagree and whether the link works. A person who leaves may have a different reason, so do not infer motive from analytics alone. If a page asks for an email address, explain what the person receives and why that information is needed. Nielsen Norman Group's work on levels of commitment argues that a site should meet basic trust needs before requesting greater personal information or commitment (Sherwin, 2016). Apply that as a design question, not as a rigid count of screens a visitor must see.
A08-05-site-trust-factors
Figure 5
Figure 5 in words: Look for usable design, up-front disclosure, current and complete answers, and verifiable connections beyond the page. These are the four categories discussed by Harley (2016), redrawn as an audit checklist. They are signs to inspect, not a score that predicts trust.

Reviews, testimonials and social proof

Social proof is information about what other people have done or experienced. Reviews, ratings and customer quotes can help a person ask whether an option suits them. They can also mislead if they are fake, selectively shown, stripped of context or presented as proof of a general outcome. A popular choice is not automatically the right choice for a particular reader.

The current CAP Code prohibits fake consumer reviews in marketing communications and requires disclosure where reviews are incentivised. It also warns against misleading presentation of review information, including suppressing negative reviews while publishing positive ones. Testimonials need to be genuine, relevant to the advertised product and used with permission where required; factual claims within them still must not mislead (CAP, n.d.). These are specific rules for communications within the Code's scope. When in doubt about a live campaign, check the current rules and your evidence.

For an honest review display, preserve the actual words and relevant context. State if a reviewer received a free product, discount or other incentive. Explain what a rating summarises and how many eligible reviews were included if you present an aggregate. Do not imply that a response from one person is representative of everyone. If a quote mentions a result, verify that it occurred and determine whether the wider ad implies a general result you cannot substantiate.

An endorsement from a person with expertise may help with a technical question, but credentials and scope matter. A professional's approval of one aspect does not automatically cover every feature. A company logo does not become a partnership badge without authorisation. “As seen in” should refer to an actual, relevant appearance that a reader could check. The business should know which organisation granted permission, for what use and for how long.
A08-06-review-provenance
Figure 6
Figure 6 in words: Begin with an actual experience, check permission and relevance, keep the context and incentive visible, and review the final display. A testimonial is a person's account, not a substitute for evidence of every factual claim on the page. The sequence is an original teaching checklist informed by CAP's review and testimonial rules.

Expertise is a clue, not proof of everything

An expert source can help you understand a question and test an argument. It should still be evaluated. Ask whether the person has relevant expertise, whether the source is independent of the seller, whether the claim is within their field, whether the evidence is visible and whether the material is current for the decision.

These distinctions matter when building a beginner article too. A textbook may teach a framework, a professional body may state standards, a regulator may explain obligations, a vendor may document its own product, and an independent study may test an outcome. Each can be valuable for a different purpose. None makes every sentence on your page true. Make the source's role clear to the reader.

If an accessible external explanation already covers a term, link to it. Add an academic or authoritative source for deeper reading when it genuinely helps. Do not pretend an unpublished chapter in your own database is a public resource. For this article, Nielsen Norman Group's explanation of website trustworthiness is a readable introduction; the CAP Code's misleading advertising section is a direct source for the UK ad claims and testimonials discussed here. Both links are optional to understanding the written lesson.
A08-07-source-roles-table
Figure 7
Figure 7 in words: Product documentation can confirm a current feature, a customer review can describe that person's experience, a study can support a finding within its design, and official guidance can state relevant rules. None alone proves a different claim outside its scope. The last column reminds you to ask what extra check is needed.

Let people recognise an advertisement

Trust also depends on knowing who is speaking. A brand's own offer page is usually recognisable as commercial. A creator's recommendation or an editorial-looking post may be less obvious if there is a paid relationship or another commercial arrangement. A reader should be able to recognise an ad before relying on it as an independent opinion.

The CAP's current advice on social media and influencer marketing says advertising should be obviously identifiable, with the commercial intent clear before or as the person encounters the content. A disclosure buried after a “See more” expansion or visible only after someone starts watching may not do that job (CAP, 2026a). The detail depends on the medium and arrangement. Do not assume that a platform's tool or a generic “thanks” line necessarily solves it; check the current guidance for the campaign.

For your own work, record who created the material, who paid for it, what was supplied and how that relationship will be made clear on the device people actually use. Review the headline, thumbnail, first screen and caption together. If a recommendation would be understood differently once the relationship is known, the relationship matters to the decision. This is a transparency check, not a reason to avoid genuine partnerships.

When urgency is real, explain it. When it is not, leave it out

Scarcity and deadlines can be genuine information. A limited number of appointments, an actual end date or an inventory constraint can matter to a decision. But a countdown that resets, an unverified “only one left” label or a comparison price with no real basis can put unfair pressure on a reader. The CMA's guidance on online urgency and price-reduction claims addresses misleading or unfair pressure from these practices (CMA, 2023).

Ask four questions before displaying a time or scarcity cue: What is the factual basis? Who checks it? What happens when it expires? Would a reader understand the limitation correctly? If you cannot answer, remove or narrow the cue. Give a person enough time to read important terms. Do not make an opt-out harder than an opt-in simply because it raises a short-term metric.

Price clarity is part of the same decision. Current CMA guidance says mandatory fees, taxes and charges should be included in the total price up front, and explains the problem of adding unavoidable charges late in a transaction (CMA, 2026). This chapter is a learning guide, not a checklist for every pricing rule in every sector. For a live offer, consult the current official guidance and check the specific terms.

Choice architecture includes the ordering, labels, defaults and prominence of options. A useful design can make a complex choice easier. A deceptive design can disguise a paid extra, hide the decline path or make cancellation needlessly difficult. The CMA (2022) and OECD (2022) discuss such harmful patterns. Use the same scrutiny on a phone screen: a disclosure that is visible on desktop may be hidden below a button or truncated in a mobile view.
A08-08-choice-design-table
Figure 8
Figure 8 in words: Unsupported countdowns call for a verified deadline or no timer. Hidden mandatory fees call for the complete price early. A preselected paid add-on calls for a clear, voluntary choice. A difficult decline route calls for a visible path to say no. These are prompts for a review, not assertions that a particular page contains these patterns.

A clear call to action still belongs on the page

A call to action is the next step a page invites the reader to take. It can be helpful: “Read the terms,” “Compare options,” “Ask a question,” “Start an application” or “Buy” are all different levels of commitment. The label should match what happens after the click. Do not use a harmless-sounding label that unexpectedly commits someone to payment or data collection.

Explain the requirements close to the action. If the person needs an account, a document, a payment card or another person's approval, say so at the appropriate moment. If the action starts a subscription, show that fact and the renewal terms clearly. A confirmation screen should tell the person what happened and how to correct a mistake. If they choose to wait or decline, the page should not punish them with confusing navigation.

The practical test is simple: Can someone who has not built the page tell you what will happen after each button, what it will cost, what data is required and how to stop? Observe them without guiding their answer. Their difficulty is evidence about the interface; it is not a judgment of their ability. Make the necessary details clearer, and check again.

Exercise: audit one claim and one choice

Allow 30–45 minutes. Use a real page or draft you are authorised to inspect. If you do not have one, use a blank template with placeholders such as [claim], [source], [date] and [limit]. Do not create a fictional customer, a made-up result or a fake review to fill the worksheet. Keep sensitive records out of the public article.

  1. Choose one decision question. Write what a reader needs to know before a particular action. State the action precisely, including the option to wait or decline.
  2. Capture one claim in context. Record the exact words, image, button and adjacent qualification. If the page is live, note its URL and access date. Ask what a new reader is likely to infer.
  3. Classify the claim. Is it a checkable feature, benefit, price, comparison, future promise, opinion or personal account? There may be more than one class.
  4. Record the proof. Name the source and owner. Check the method, date, relevant population or conditions, and what remains uncertain. Write “no suitable evidence yet” where that is true.
  5. Check the limit. Does the claim apply to all uses, or only some? Is a necessary condition visible with it? Does the qualification merely clarify, or does it contradict the headline?
  6. Inspect the decision path. Find the total price, next step, information request, ability to decline and route to support. View the same page on a phone-sized screen.
  7. Revise one thing. Narrow an unsupported assertion, add a material condition, make a cost visible or relabel a button to match its action. Do not replace one ambiguity with another.
  8. Test comprehension. Ask a reader what they think the revised claim and button mean. Record their words and any uncertainty. If no one is available, mark the test not done and keep it as the next check.
A08-09-claim-audit-worksheet
Figure 9
Figure 9 in words: Write the exact claim and the interpretation a reader may take from it. Add a source and date, an important limit, an owner and the next check. The worksheet deliberately contains no invented evidence or quote. A blank or “unknown” is an honest entry.

If you find a claim that cannot currently be supported, the best revision may be to delete it. If you find a page that asks for payment before explaining the total, the necessary action is to fix the information and flow. A better headline alone will not repair the choice.

Review before and after publication

Marketing pages change. A price is updated, a feature is renamed, a review is removed, a mobile layout breaks or a timed offer expires. Set a practical review interval for claims that can change, and identify who owns each source. Keep a version of the page or message as published, with the supporting record and date. If evidence no longer fits, edit or withdraw the claim promptly.

Measure more than the number of clicks. An increased click rate could coexist with greater confusion or more unsuitable enquiries. Look for comprehension, complete information, relevant questions, successful task completion and complaints. When you alter a page, note what else changed at the same time; a before-and-after count alone cannot prove cause. If a clear limitation reduces the number of unsuitable sign-ups, that may be a better outcome than a higher raw conversion figure.

A08-10-review-cycle
Figure 10
Figure 10 in words: Start with the reader's question, draft a narrow claim, check evidence and limits, present a fair choice, observe understanding and revise. The final review returns to the question because the offer, audience and available information can change. This is a quality cycle, not a promise of increased sales.

Optional videos

These two YouTube videos give different ways to think about trust and influence. Their original-source pages, public descriptions and institutional provenance were checked for this draft; the videos were not played in full, and captions or regional playback were not verified. Check them before publication. Neither replaces the written explanation or the current UK sources above.

Short visual overview: 4 Trustworthiness Factors, Nielsen Norman Group / Aurora Harley (2020), approximately three minutes according to the publisher's video listing. Watch for the four website factors, then compare them with Figure 5. This is professional usability teaching, not a test of your own page.

Deeper persuasion lens: Science Of Persuasion, Robert Cialdini and Steve Martin / Team Cialdini (2012), about 11:50 in a course listing. Arizona State University identifies Cialdini as a professor emeritus associated with this educational video. It introduces principles of influence. The ethical task is to ask whether a cue describes something true and relevant or merely pushes a person to act. The video is an academic researcher's educational explanation, not evidence that a particular business claim is true.

What to remember

  • Help the reader answer a genuine decision question. A confident “no” can be a valid result.
  • Check the claim as the whole page presents it, including images, context and what a reader is likely to infer.
  • Match objective statements to documentary evidence. Put important limits beside the claim.
  • Genuine reviews are accounts of experience. They do not by themselves prove a universal outcome.
  • Show the real price and conditions, use real deadlines, and make the next step and the decline path understandable.
  • Keep a dated claim record. Revisit it when the offer or its evidence changes.
Next in the series: Chapter 9, Reach relevant people using behaviour, location and interests, will examine when targeting is useful and appropriate. Add its canonical public link only after it is published and verified.

References

Committee of Advertising Practice (CAP) (n.d.) ‘03 Misleading advertising’, The UK Code of Non-broadcast Advertising and Direct & Promotional Marketing. Available at: https://www.asa.org.uk/type/non_broadcast/code_section/03.html (Accessed: 14 November 2026).
Committee of Advertising Practice (CAP) (2026a) ‘Recognising ads: Social media and influencer marketing’, AdviceOnline. Available at: https://www.asa.org.uk/advice-online/recognising-ads-social-media.html (Accessed: 14 November 2026).
Committee of Advertising Practice (CAP) (2026b) ‘Substantiation’, AdviceOnline, 21 September. Available at: https://www.asa.org.uk/advice-online/substantiation.html (Accessed: 14 November 2026).
Competition and Markets Authority (CMA) (2022) Online Choice Architecture: How digital design can harm competition and consumers. 5 April. Available at: https://www.gov.uk/government/publications/online-choice-architecture-how-digital-design-can-harm-competition-and-consumers (Accessed: 14 November 2026).
Competition and Markets Authority (CMA) (2023) Using urgency and price reduction claims online. 29 March. Available at: https://www.gov.uk/government/publications/using-urgency-and-price-reduction-claims-online (Accessed: 14 November 2026).
Competition and Markets Authority (CMA) (2026) ‘Providing clear and accurate information about prices: summary’, GOV.UK, updated 7 January. Available at: https://www.gov.uk/government/publications/price-transparency-cma209/providing-clear-and-accurate-information-about-prices-summary (Accessed: 14 November 2026).
Harley, A. (2016) ‘Trustworthiness in web design: 4 credibility factors’, Nielsen Norman Group, 8 May. Available at: https://www.nngroup.com/articles/trustworthy-design/ (Accessed: 14 November 2026).
Organisation for Economic Co-operation and Development (OECD) (2022) Dark commercial patterns. OECD Digital Economy Papers, no. 336. Paris: OECD Publishing. Available at: https://www.oecd.org/content/dam/oecd/en/publications/reports/2022/10/dark-commercial-patterns_9f6169cd/44f5e846-en.pdf (Accessed: 14 November 2026).
Sherwin, K. (2016) ‘Hierarchy of trust: The 5 experiential levels of commitment’, Nielsen Norman Group, 6 March. Available at: https://www.nngroup.com/articles/commitment-levels/ (Accessed: 14 November 2026).

Supplementary videos: Harley, A. (2020) ‘4 Trustworthiness Factors’, YouTube, Nielsen Norman Group, 30 October. Available at: https://www.youtube.com/watch?v=pAWcYSBQHNc (Accessed: 14 November 2026). Cialdini, R. and Martin, S. (2012) ‘Science Of Persuasion’, YouTube, Team Cialdini. Available at: https://www.youtube.com/watch?v=cFdCzN7RYbw (Accessed: 14 November 2026).