Key Takeaways
- Test whether people understand your offer, not whether they like your favourite headline.
- Choose participants who resemble the customers you want. Colleagues and friends already know too much about the business.
- Keep the test simple: show a realistic page, listen to the explanation and ask what information is missing.
- Record misunderstandings and unanswered questions, not just positive comments. Small qualitative tests do not establish market-wide percentages.
- Use the findings to change the brief before investing in detailed design, photography and development.
The best time to discover that customers misunderstand your brand message is before the website redesign, while changing the words is still easy. A short, well-run conversation can reveal that your impressive headline means something quite different to the person you want to reach.
You do not need to ask people to become copywriters. You need to find out what they believe you sell, whether it sounds relevant to their situation, which claims they trust and what stops them taking the next step.
This guide sets out a practical message-testing exercise for a small service business. It is a way to improve decisions, not a promise of statistically proven conversion growth. If the decision is high-value or the audience is complex, specialist research may be appropriate. The principle stays the same: collect evidence before defending a design.
Decide What You Need to Learn
Start with a question you could answer badly. “Do people like our brand?” is too broad. “Do office managers understand that we manage the complete refurbishment rather than only supply furniture?” gives the session a useful purpose.
Write a short intended-message statement in private. Include the customer, the problem, the service, one meaningful difference and the next action. This becomes the reference against which you compare what participants actually understood. Do not show it to them first.
For a fictional website support business, the intended message might be: “We maintain established WordPress websites for small professional firms, with a named contact and a clearly defined support plan.” That is more testable than “We enable digital excellence”. It also helps the designer decide which evidence belongs near the headline.
If the team cannot agree on this statement, the problem comes before copy testing. A focused brand strategy exercise may be needed to settle the offer and audience. Our explanation of a brand strategy workshop covers the broader decisions; this article is about checking how one message lands.
Recruit People Who Can Give Relevant Answers
Choose actual or plausible buyers. Include people who recognise the problem but have not memorised your website. A current customer may help explain buying concerns, while a prospective customer can expose language that only makes sense after working with you.
Do not treat a handful of conversations as a representative survey. Use them to find concrete misunderstandings and generate changes worth testing again. If your business serves different audiences, keep their responses separate. A finance director and a technical manager may need different evidence even when they buy the same service.
Explain what the session involves, how notes or recordings will be used and whether you will keep identifying details. Ask permission before recording. Offer a reasonable way to decline or stop, and avoid collecting personal information that is irrelevant to the exercise.
The GOV.UK guide to in-depth interviews is a useful reference for planning purposeful conversations and preparing a discussion guide. Your task here is narrower: understand the participant's interpretation of a commercial message without selling it to them during the session.
Show Something Realistic, but Keep It Cheap to Change
A plain text document can test language, but a simple page layout provides context. People interpret a headline alongside the navigation, images, service labels and call to action. Use a basic prototype with enough structure to resemble the eventual page.
Keep the design restrained if your main question is about wording. Otherwise, an attractive photograph or bright colour may dominate the conversation. If you compare two message versions, hold the rest of the layout reasonably consistent so you can understand what changed.
Do not put one option in a polished mock-up and the other in a rough paragraph. That tests production quality as much as the message. Also avoid labels such as “new improved version” or “our preferred option”. Participants should not have to guess which answer will please the person asking.
Show the page on a device that suits the task. If many enquiries start on phones, include a phone view rather than relying entirely on a large desktop presentation. The UX and UI design process should connect message decisions with what people can actually see and use.
Use a Neutral Conversation Script
Open by saying that you are testing the page, not the participant. Make it clear that confusion is useful feedback. Ask them to read the page as they normally would, rather than instructing them to study every sentence.
Then work through a short script:
- What do you think this business does?
- Who do you think this service is intended for?
- What problem would make you consider contacting them?
- What seems different from another supplier you might use?
- Which statement would you want evidence for?
- What would you expect to happen after pressing the main button?
- What would you need to know before deciding whether to enquire?
These are prompts, not a rigid questionnaire. Follow an interesting answer with “What makes you say that?” or “Which part of the page gave you that impression?” The goal is to trace an interpretation back to the wording or layout.
Avoid “Was it clear that we include ongoing support?” That reveals the answer. Ask “What do you think happens after launch?” instead. Resist filling an awkward silence. If you explain the page before the participant answers, you have tested your explanation rather than the page.
Separate Understanding from Preference
People may prefer a short, stylish headline while misunderstanding the service. Others may call the page boring but accurately describe the offer and feel confident about contacting you. Neither response should be reduced to a simple like or dislike.
Record these dimensions separately:
| Dimension | What to listen for | Useful evidence |
|---|---|---|
| Comprehension | Can they describe the service accurately? | Their own explanation |
| Relevance | Can they recognise a situation where it would help? | A specific problem or occasion |
| Difference | Do they identify a meaningful reason to choose it? | A comparison in their own words |
| Trust | What needs proof or qualification? | A claim they question |
| Next step | Do they understand the action and its consequences? | Their expectation of the enquiry process |
This is a decision sheet, not a personality test. Do not assign a made-up scientific score or claim that four positive responses prove market demand. Keep the original observation beside any summary so colleagues can see what the conclusion rests on.
Look for the Wrong Assumptions
The most valuable finding may be an assumption nobody in the business noticed. A phrase such as “complete website support” could mean unlimited edits, hosting, security monitoring or telephone help, depending on the reader.
Consider a fictional test of a maintenance service. One participant assumes the plan includes a redesign. Another expects same-day content changes. A third cannot tell whether the business supports sites built by other developers. Those are different problems, and a louder headline will not solve them.
The response might be a clearer scope line, a short inclusions table and an explicit answer about existing websites. It might also reveal an offer problem: perhaps the business genuinely has not decided what the plan includes. Research should be allowed to change the service brief, not only the adjectives.
Our guide to brand positioning explains how difference relates to the market. In this exercise, ask whether that difference is visible and believable to the person buying. If nobody can identify it without a lengthy explanation, the website has more work to do.
Test an Action as Well as a Headline
After discussing the message, give the participant a realistic task. For example: “You already have a website and want someone to maintain it. Find out what you would need to send before asking for a quote.”
Observe where they go and what they expect. Do not immediately guide them to the right button. This can expose a gap between understandable copy and usable navigation.
The GOV.UK moderated usability testing guide describes observing people attempt tasks using a service. Applied to your redesign, the useful distinction is between what somebody says they would do and the route they actually take through the page.
A button labelled “Let's talk” may sound friendly but leave the next step unclear. A more specific label or a short explanation beside the form may answer the concern. Test that revision rather than assuming that a more direct phrase is always better.
Turn Findings into Changes the Team Can Make
Group the observations by problem, not by participant. Separate a repeated misunderstanding from an isolated preference. Note whether a finding affects the core offer, a particular audience or a minor wording choice.
For each important issue, write four things: the observation, your interpretation, the proposed change and what would count as an improvement. For example: “Participants expected a free design concept after enquiry; the button promises too much; clarify that the first step is a discovery conversation; re-test whether people describe that step accurately.”
Keep an unresolved list. Do not force every question into a design decision during the first meeting. Pricing, delivery scope and ownership may need input from the people responsible for those commitments.
Then test the revised message with fresh participants where practical. People who saw the first version now know more, so they are not a clean test of a first impression. Retain version names and dates to avoid comparing notes from different pages by accident.
Put the Evidence into the Redesign Brief
The final output should be compact: your intended message, the main audience, tested wording, unanswered questions, proof needed and any changes to the customer journey. Include examples of misunderstandings with identifying information removed where appropriate.
This gives a website redesign team something more useful than “make it modern”. It tells them what the page must communicate and which assumptions it must avoid.
Do not treat the test as the final word on conversion. Once the website is live, review the quality of enquiries, sales questions and observed behaviour alongside analytics. A clearer message should help people make a more informed choice, including deciding that a service is not for them. Start by making that choice easier to understand; the colours and animation can then support a message that already has a purpose.
Frequently Asked Questions
How many customers should we test a brand message with?
Use a manageable first round of relevant participants to identify misunderstandings, then test revisions. The right number depends on the audiences and decision. A small qualitative round is useful for learning, but it does not produce reliable market-wide percentages.
Should we ask customers which headline they prefer?
You can ask about preference, but test understanding first. Ask what the business does, who it helps and what would happen next. An attractive headline may still give people the wrong impression of the service.
Can we test the message before the website is designed?
Yes. A simple page prototype can provide enough context while remaining cheap to change. Keep visual treatment reasonably consistent when comparing wording so that the design does not dominate the result.
Should current customers be included in message testing?
They can explain real buying concerns, but they already know your business. Include plausible buyers who are less familiar with the offer as well, and keep different audience groups visible in the findings.
What should the designer receive after the test?
Provide the intended message, tested wording, main misunderstandings, required proof and proposed changes to the customer journey. Include unresolved decisions and the evidence behind each recommendation rather than only a list of preferred phrases.
