Customer Interviews: Ask Better Questions Before Rewriting Your Website

Before asking customers which headline they prefer, ask about the decision they actually made. What started the search? What did they need to know? What happened before they felt ready to contact a provider?

Those answers can reveal information your page should explain. They can also show that the obstacle sits somewhere else.

Start with these questions:

• Tell me about the last time you looked for a provider for this work.
• What prompted you to start looking then?
• What did you do first?
• What did you need to know before contacting anyone?
• Who else became involved, and at what point?
• What happened next?

Keep the conversation grounded in that episode. “Would a checklist make you buy?” invites a prediction. “What information did you gather before asking your colleague to approve the visit?” gives you a specific process to understand.

A fictional conversation, from question to page idea

The following exchange is original and entirely hypothetical. Cedar Desk Maintenance, the buyer and their statements are invented teaching material. No customer interviews were conducted for this guide.

Cedar is considering a new section on its assessment page. The team wants to know what buyers need before requesting an initial workplace-maintenance assessment.

Interviewer: “Tell me about the last time you looked for a maintenance provider.”

Buyer: “We’d moved into another office and needed someone to look after it. I was gathering options for the operations manager.”

Interviewer: “What did you do first?”

Buyer: “I checked which companies worked in our area. Then I tried to work out what I needed before contacting them.”

Interviewer: “What were you trying to work out?”

Buyer: “The page said ‘Book an assessment.’ I didn’t know whether that meant agreeing to a contract or just having someone look at the space.”

Interviewer: “What did you do at that point?”

Buyer: “I sent the link to our operations manager and asked if I should book it. They asked what we needed to prepare, and I couldn’t answer.”

Interviewer: “What happened after that?”

Buyer: “I called the company. They explained that the visit was to understand the work and that they’d send a proposal afterward.”

Interviewer: “Was anything still unresolved?”

Buyer: “I needed to know whether our facilities colleague should be there and whether we should have floor plans ready.”

The page idea follows from the questions in the account: explain what the assessment involves, what commitment it creates and what information is useful beforehand. The service owner must supply the real answers. The researcher should not invent “free,” “no obligation” or a response deadline because those phrases sound reassuring.

This conversation does not show that the buyer failed to use the website, or that a new section would increase enquiries. It shows how a specific buying account can lead to a specific content question.

Choose the buying question before interview questions

A useful research question names something the team does not understand. “What information do first-time buyers need before requesting an assessment?” is a workable example. “How can we persuade customers that our new page is better?” starts with a conclusion.

Separate your proposed change from the uncertainty behind it. You may want to add a pricing table, but the real uncertainty might be whether buyers understand the scope of a quote. Interviews should leave room for an answer that changes your plan.

Write down the page or journey in scope and the decision the findings will inform. An assessment page, a software-demo page and a service-renewal email involve different buying situations. A broad conversation about the company can produce interesting material while missing the decision you need to make now.

Include a competing explanation. Cedar’s buyers might hesitate because the assessment is unclear. They might also understand it fully but need internal budget approval. Those explanations lead to different actions.

Decide what would change your mind. If people consistently describe clear assessment information but delays caused by an unavailable approver, a page rewrite may have limited value. If they repeatedly need the same practical explanation before contacting anyone, the team has a content question worth investigating.

The GDS guidance on in-depth interviews recommends open questions and concrete accounts. Here, that approach serves a narrow marketing task: learn enough about a buying decision to improve the information a page provides.

Interview the decision, not the website

Customers can tell you about the situation they faced, the alternatives they considered and the information they sought. They are not responsible for designing your page.

Asking “What should our website include?” can produce a feature wish list. Asking “What did you need to know before making a recommendation to your manager?” reveals the work the person was trying to complete.

Start before the website entered the story. A workplace move, an expiring contract or a recurring service problem may explain why the buyer searched at that particular moment. Without that context, a page team may treat every enquiry as if it arose from the same need.

Follow the sequence. What happened before the first call? When did the buyer involve a colleague? Did they compare written proposals before or after a site visit? A statement about price means something different depending on what the person knew at the time.

Keep reported experience distinct from observed use. Hearing “I couldn’t find the service area” is useful interview material. Watching someone attempt to find it on the website is a different kind of evidence. If you need the latter, use website usability testing.

You can combine the methods in one session if planned appropriately, but label the activities and conclusions separately. Discussing a past purchase does not establish that a person can complete the current form. Completing a form does not explain the full buying process.

Replace questions that point to your preferred answer

Leading questions often sound friendly. They still introduce an explanation, emotion or solution before the participant has supplied one.

Follow-ups can stay simple. “What happened after that?” helps reconstruct the episode. “What did that word mean to you?” checks terminology. “Can you describe a specific example?” moves a general statement toward something you can interpret.

Ask one question at a time. “How did you find us, compare us, decide on budget and book?” is difficult to answer clearly. It also makes it easy for the interviewer to hear only the part that supports their plan.

Do not supply categories too early. A buyer may care about access to the premises, disruption to staff or the ability to explain the proposal to a board. A forced choice between price, speed and quality can hide those concerns.

If you later probe a topic that did not arise naturally, keep that context. An answer to “Did you look for insurance information?” is different from an unprompted account of searching for it. Both can be useful; they should not be described as identical evidence of salience.

Build a conversation around a recent episode

You need a discussion guide, but it should support listening rather than become a questionnaire read at speed. For Cedar, a practical flow has five parts: the situation, the search, the comparison, the internal decision and the next step.

Understand the situation

“Tell me about the office and your role in arranging the work.”

“What changed that made you look for a provider?”

“Had you arranged something similar before?”

These questions establish relevant experience. An experienced facilities manager renewing a contract may need different information from an office coordinator arranging maintenance for the first time.

Reconstruct the search

“What did you do first?”

“Where did you look for information?”

“What made a company worth considering further?”

Let the person describe channels in their own terms. Do not assume the website was the only influence. A colleague’s recommendation, a previous provider or an existing contract may shape the shortlist before any landing page is opened.

Explore the comparison

“What did you need to compare?”

“Was anything difficult to establish?”

“What information did you ask a provider for?”

When someone says one option was “better,” ask what that meant in the situation. It could refer to scope, availability, communication, procurement requirements or another factor. Resist converting a broad adjective into a headline immediately.

Understand the internal decision

“Who else became involved?”

“What did they need from you?”

“What had to happen before you could proceed?”

Avoid asking for colleagues’ names or confidential documents when roles and process are enough. The useful question is how the decision moved, not the private details of everyone involved.

Follow the outcome

“What happened after you contacted the provider?”

“What, if anything, delayed the next step?”

“Looking back at that particular decision, what information was missing when you needed it?”

Close by summarizing your understanding and inviting correction. If the buyer says you combined two different issues, keep the correction. It may be the most useful part of the conversation.

Recruit people who can describe the relevant decision

Your most enthusiastic customer is not automatically the best person to ask about an initial enquiry. A long-standing account may have forgotten the first purchase or now bypass the website entirely.

Look for people who recently faced the situation in scope. For Cedar, that could include someone who requested an assessment, someone who postponed and someone who chose another route. The experience should be relevant even when the outcome differs.

Ask about their role and broad outcome, rather than collecting unnecessary personal or commercial details. “Did you help compare providers or approve the work?” can establish fit. You do not need the competing proposals or the organization’s confidential budget to begin.

Include useful contrasts. First-time and experienced buyers may use the same words differently. A person preparing a recommendation may need something shareable; the approver may need commercial terms. Someone who stopped before contacting sales may reveal a gap missing from successful-customer accounts.

Be honest about recruitment limits. A customer newsletter reaches people already connected to the business. A sales shortlist may favour strong relationships. People willing to join a video call may differ from those who prefer another format. Those gaps belong in the interpretation.

A conversation with a former prospect should remain voluntary and clearly separated from prospecting. Do not disguise a sales follow-up as research. If the person asks for commercial help during the interview, agree a separate follow-up rather than turning the research into a pitch.

Offer a workable format and language. Ask about access needs without demanding unnecessary medical details. Make scheduling, compensation and the ability to decline straightforward. The quality of an interview depends partly on whether the person feels free to disagree.

Plan enough interviews to challenge the explanation

There is no universal interview count that makes every page decision well supported. A narrow question about one service and a similar buying situation differs from a study spanning several products, languages and approval processes.

Set an initial round you can recruit, conduct and analyze properly. At a review point, ask what you have learned and what important situations remain missing. If every participant came through the same sales relationship, repeated answers may reflect that recruitment route.

The UK government’s qualitative-interview guidance explains that interviews can explore a range of views without establishing how common each view is. Use that distinction when presenting findings to the team.

Several passages from one person are still one person’s account. Two people from the same buying committee may describe the same purchase episode. Neither should be counted as a set of independent confirmations just because the notes occupy several rows.

Look for information that could challenge the proposed explanation. If you think unclear commitment deters assessment requests, include people who understood the commitment yet delayed. They may reveal an approval, timing or service-fit issue that the page cannot resolve.

If time or budget ends the research, say so. A bounded decision with visible gaps is more credible than declaring the whole audience understood because the team has run out of sessions.

Keep the invitation and permission clear

Tell participants who is conducting the research, why, what the conversation involves and how long you expect it to take. Explain that they can skip questions or stop and that declining does not affect the service they receive.

Before recording or using automated transcription, explain the collection, its purpose, who receives it and how it will be handled. A notes-only route may be appropriate when the team can support it. Do not assume that agreeing to an interview includes every recording or sharing method.

The Office of the Privacy Commissioner of Canada’s meaningful-consent guidance emphasizes understandable information about collection, purposes, recipients and relevant risks. The organization must determine which obligations apply to its circumstances; an interview template does not provide a legal sign-off.

Decide access and retention before collecting records. The writer considering an FAQ may need a finding and a short supporting excerpt, not an identifiable recording. Keep scheduling information separate from interview notes where practical.

Explain withdrawal honestly. Provide a contact and describe what can be stopped or removed, including any applicable limits. Avoid making an unconditional deletion promise the organization cannot carry out.

Research permission is also separate from permission to publish a quote or use someone as a testimonial. A phrase that would make appealing website copy still requires an appropriate public-use decision. Often the better response is to write an original explanation of the service.

Listen without defending the company

The interviewer’s job is to understand the account. Correcting every misunderstanding or explaining the intended website design can replace the customer’s experience with the company’s preferred story.

If someone says “I thought the assessment meant signing a contract,” ask what led them to that interpretation and what they did next. You can clarify the actual offer later where appropriate, but first understand the sequence that matters to the research.

If someone says the website was clear, accept that as a possible account. Ask them to describe how they reached the next step. Do not keep asking the same question until they identify a problem.

Allow pauses. People may be remembering a sequence rather than withholding an answer. Filling every silence with examples can supply categories they had not considered.

Show interest without rewarding agreement. “That’s exactly what we thought” signals that one answer is welcome. Neutral follow-up gives contrasting accounts room to appear.

Think carefully about observers. A customer may be less candid with their account manager watching. Disclose who is present and consider whether that person needs to attend. A silent observer can still affect the conversation.

If the account moves into confidential or sensitive material outside the research purpose, redirect it. Ask about the process at a general level rather than collecting another organization’s private details. Follow the agreed handling procedure for anything accidentally disclosed.

Resolve ambiguous answers before changing the page

Customer language is useful because it shows how people frame a problem. It still needs interpretation.

Suppose an invented Cedar buyer says, “The assessment felt expensive.” That could mean the fee was high, the time commitment seemed large, or they feared an obligation after the visit. Asking “How much cheaper should it be?” commits to one interpretation too soon.

Better follow-ups are “What made it feel expensive?” and “What did you understand was included?” Then ask what they compared it with and what they did next. The resulting account may call for price clarity, scope explanation, a different offer or no page change at all.

Similarly, “I wanted something quicker” might mean a shorter form, a faster appointment or a decision that needed fewer internal approvals. The same word can point to different parts of the experience.

Do not treat a suggested solution as a complete diagnosis. If a buyer asks for a chatbot, ask what they wanted to find out and how they tried to do it. A clear explanation on the page might answer the need; in another situation, a conversation really may be required.

Keep the original phrasing and context in the notes, then write your interpretation separately. That allows another reviewer to challenge the conclusion without losing what the person actually said.

Decide when to show the page

If the aim is to understand a recent buying episode, begin with the account before showing a proposed redesign. A new page can introduce language and explanations that were absent when the person made the decision.

For Cedar, a buyer might initially describe uncertainty about whether an assessment meant a contract. If the interviewer first presents a large “No ongoing commitment” headline, the subsequent conversation has been influenced by that explanation. The reaction to the draft may still be useful, but it answers a different question.

Keep the activities in sequence and label the transition. “We’ve talked about what happened at the time. I’d now like to show you a draft and hear what it communicates.” That helps both the participant and the note-taker distinguish recalled experience from a current response.

Ask what the draft means before asking whether they like it. “What would you expect to happen after selecting this?” is more useful for a next-step explanation than “Is this reassuring?” If they understand it differently from the intended process, ask which wording produced that interpretation.

Do not expect a participant to predict the draft’s commercial impact. They can explain their reaction and identify questions. Saying they would contact the company after seeing the page is a stated intention, not an observed enquiry or future sale.

Use sales knowledge as a start, then examine it

Sales and service teams often know the questions people ask repeatedly. Their notes can help choose a research question and identify language worth exploring. They also reflect the purpose and perspective of the original conversation.

Suppose sales says, “People always ask about price.” Ask when the question appears and what the buyer is trying to establish. Someone comparing complete proposals may ask a different pricing question from someone trying to understand whether the first assessment has a fee.

Interviewing can add the missing context. The useful finding might concern price visibility, but it might instead concern uncertainty about scope or commitment. A page that publishes a number without explaining the service could leave the original uncertainty intact.

Review authorized sales notes appropriately, and keep their origin visible. A salesperson’s summary is not a verbatim customer quotation. A field labelled “lost on price” may compress a more complicated decision. Do not relabel those records as interviews merely because they support the same theme.

When sales colleagues contribute observations, ask them to separate what the buyer said from their interpretation. That makes their experience more useful and reduces the chance that research becomes a contest between teams defending different stories.

When several people describe one purchase

B2B decisions often involve more than one person. Speaking with both the person gathering information and the approver can reveal a handoff that a single account misses.

Keep their roles and the shared episode connected. The coordinator may say they needed a simple explanation to forward. The approver may say they needed to know whether the assessment created an obligation. Those are complementary views of one decision, not proof that two separate customers faced the same issue.

Ask each person about the part they actually experienced. An approver may not know which website pages the coordinator read. The coordinator may not know every reason behind a budget decision. Avoid using one person’s account as certainty about another person’s motives.

A joint interview can reveal how people work together, but it can also make disagreement harder to express. Choose the format according to the question and the relationship, explain who will attend and give participants a practical choice. Where separate conversations are more appropriate, compare them carefully without treating differences as dishonesty.

For page planning, the resulting insight may be a need for an explanation that travels well between roles. That could be a concise scope summary or clear next-step section. Verify the need before creating an entire library of sales materials the research did not call for.

Turn the opening conversation into a page section

Return to Cedar’s fictional buyer. They reported two questions before requesting an assessment: whether the visit created a commitment and what they should prepare with their colleague.

A reasonable content proposal is a compact section titled “Before you request an assessment.” Its job is to answer those questions beside the enquiry route.

The team needs factual answers from service delivery before writing it. What happens during the assessment? Is there a fee? Does requesting or attending create any obligation? What information is useful beforehand? Who should attend? What happens afterward?

For teaching purposes, suppose the fictional service owner confirms these terms: the assessment is a scoping conversation and site review, it does not itself commit the business to ongoing maintenance, and a separate proposal follows. Preparation requirements depend on the premises. These assumptions apply only to this example.

A draft passage could then read:

“The assessment helps us understand your workplace and the maintenance you need. We’ll review the space and discuss the work before preparing a separate proposal. The assessment itself does not commit you to an ongoing maintenance contract. Before the visit, we’ll confirm which details would help and who should attend.”

The fee still needs an explicit, verified statement in a real version. The hypothetical assumptions above do not establish one, so the writer should not add “free.” This is an example of how research can expose a missing answer without authorizing the writer to supply it.

Place the answer where the question matters

A general FAQ at the bottom of the website may be less useful than a short explanation near “Request an assessment.” People need the information when deciding whether that step is appropriate.

Keep the main answer visible and concise. Supporting detail can sit beneath it if the preparation varies. Avoid burying the commitment explanation inside a long paragraph about company values.

Use the customer’s underlying question to choose the heading. “What happens during the assessment?” is direct. “Our discovery methodology” may reproduce the internal terminology that caused the uncertainty.

Preserve the operational truth

If some assessments have a fee or special conditions, the page must explain those accurately. A reassuring phrase should not erase exceptions that change the decision.

The same applies to preparation. Do not promise that “no preparation is needed” merely because one participant found the process demanding. Perhaps floor plans are optional, or perhaps a particular technical assessment genuinely requires them. Ask the person delivering the service.

Research informs what to explain. Operations establishes what the explanation can truthfully say.

A negative case: the page cannot approve the budget

Now consider another fictional Cedar interview. This buyer understood the assessment, knew what to prepare and wanted to proceed. They postponed because the budget owner was away and no one else could authorize the work.

That account challenges the broad claim that clearer assessment copy will remove every delay. Rewriting the button cannot grant spending authority.

The page might still help the buyer prepare a recommendation if they need a concise scope summary or a clear explanation of next steps. But that is a different question. Ask what the approver needed and whether the buyer had it, rather than assuming that another website section will solve the problem.

The appropriate response may be to leave the page alone and improve a sales handoff, or simply recognize that the purchase has an external timing constraint. Research has still been useful: it prevented a misleading diagnosis.

Keep this account beside the clarification findings. Do not remove it because it complicates the story. A good synthesis explains where a proposed change may help and where it is unlikely to matter.

The same logic applies to other constraints. A provider cannot use copy to expand a service area it does not cover, create appointment capacity it does not have or change a buyer’s procurement policy. The website can communicate those realities clearly while the business decides whether operational change is warranted.

Compare accounts without flattening the differences

After each conversation, preserve the important notes before discussing solutions. Record the prompt, the relevant account and whether the wording is an exact quote or a paraphrase. Use participant codes in working summaries where appropriate.

A direct quote needs a reliable source and permission for the use. If you captured only a summary, keep it as a summary. Do not polish it into quotation marks because a vivid sentence would strengthen the presentation.

Group related questions in plain language: assessment commitment, preparation, approval, service fit or timing. You do not need an elaborate coding scheme for a small page decision, but you do need enough consistency to compare like with like.

The GDS analysis guidance distinguishes observations, findings and actions. Apply that separation to the account, your explanation and the proposed content response.

Watch for differences hidden under the same theme. “Preparation” for an office coordinator may mean gathering information from facilities. For an experienced manager it may mean arranging access for the visit. Those needs could require different wording or placement.

Keep missing information distinct from contrary information. A participant who never mentioned price has not established that price was irrelevant. They may not have been asked, may not remember or may have focused on another part of the decision.

Review interpretations with someone willing to challenge them. Ask what else could explain the account and which evidence would contradict the recommendation. The most articulate participant should not automatically define the entire website strategy.

Write a recommendation the team can verify

A useful recommendation connects a buying situation to a specific question and a proportionate response.

For Cedar, it could be:

“First-time buyers preparing an internal recommendation may need a clearer explanation of assessment commitment and preparation. Add that information beside the request route after service delivery confirms the terms. The fictional approval-delay account illustrates why the section would not resolve every postponed enquiry.”

In real work, support the finding with the relevant session notes and describe recruitment limits. Keep the report audience’s access appropriate to the underlying material.

Avoid vague tickets such as “Add trust” or “Make the page customer-focused.” Name the question to answer, where the answer belongs and who can verify it. A writer can then develop useful content rather than guessing what the research was intended to change.

The recommendation may be to remove an unnecessary explanation, move existing information or leave the page unchanged. Interviews do not need to produce more copy to be worthwhile.

For broader inspection of the current page, use the landing-page audit checklist. The interview findings provide buying context; the audit examines the actual page and route. Together they can inform a focused decision without making either method claim more than it establishes.

Check understanding before claiming business impact

First verify the draft’s factual statements. A clear explanation of the wrong offer is still a problem. Service, pricing and commercial owners should confirm the parts within their responsibility.

Then investigate the remaining uncertainty. Ask suitable people what they understand from the new section. If the question concerns whether they can find and use it, observe a realistic task. These checks can show whether the information communicates what you intend.

Commercial outcomes require a separate comparison. Define what matters before looking at results: suitable assessment requests, mistaken assumptions about fees, clarification workload or another relevant outcome. Total form submissions alone may not tell you whether the page helps the right people decide.

If traffic and conditions support an experiment, plan it appropriately. If you use a before-and-after review, account for changes in traffic, offers, staffing and season. A rise after publication does not isolate the effect of a paragraph by itself.

For a low-volume site, a research-informed improvement may remain worthwhile without a measurable conversion claim. Correcting an inaccurate statement is justified by accuracy. Explaining a recurring question is a reasoned editorial decision. Neither needs an invented percentage to sound useful.

Keep the explanation current when the service changes. The interviews describe particular decisions in a particular context. They do not make an old fee, preparation rule or next-step promise permanently true.

Use the templates to prepare a real conversation

The customer-interview templates include a prompt bank, interview plan, privacy-planning checklist and blank and fictional analysis files. Start real research from the blanks and keep the teaching examples separate.

Use the guide to choose the buying question and relevant participants. Rehearse the prompts with a colleague to catch confusing wording. That rehearsal improves the procedure; it does not count as customer evidence unless the person genuinely belongs in the planned study.

During interviews, follow the episode, leave room for disagreement and ask for clarification before reaching a conclusion. Afterward, turn the strongest supported questions into a short recommendation, with contrary accounts and factual checks visible.

The best next page change may be a clearer sentence about commitment, a practical preparation section or no change at all. Customer interviews help the team choose by understanding the decision behind the enquiry. For support connecting that research with a useful page response, explore Canada Create’s conversion-rate optimization services.

Frequently asked questions

How do you conduct customer interviews before rewriting a website?

Ask about the last time the person looked for a provider: what prompted the search, what they did first and what they needed to know before contacting anyone. Keep the conversation on that real episode rather than on opinions of your page. Follow the sequence through who else became involved and what happened next.

How many customer interviews are enough for a page decision?

There is no fixed number in the guide. Plan enough interviews to challenge your explanation, including at least one competing explanation, such as buyers understanding the page but waiting on budget approval. Decide in advance what evidence would change your mind.

Why do customer interviews before changing website copy?

They can show what information buyers needed before they felt ready to contact a provider. They can also show that the obstacle sits somewhere else, such as an approver who is unavailable. Either finding saves the team from rewriting a page for the wrong reason.

What customer interview questions work best for conversion research?

Use open questions about a recent decision: Tell me about the last time you looked for this work. What prompted you to start then? What did you need to know before contacting anyone? Who else became involved, and at what point?

Is there an example of a customer interview leading to a page change?

The guide includes a fictional exchange with a buyer for Cedar Desk Maintenance. The buyer could not tell whether booking an assessment meant agreeing to a contract or just having someone look at the space, or what to prepare. That points to a page section explaining the assessment, the commitment it creates and what to have ready, with real answers supplied by the service owner.

How is a customer interview different from usability testing?

An interview gathers reported experience: the situation, alternatives considered and information sought. Usability testing watches someone try to use the website. Hearing that a buyer could not find the service area differs from watching them search for it, and the two methods can be combined.

What are red flags in customer interview questions?

Questions that point to your preferred answer, ask for predictions or ask customers to design your page. Would a checklist make you buy? invites a guess, while asking what the person gathered before seeking approval reveals a real process. Avoid defending the company while you listen.

How should customer interview answers be handled before changing a page?

Resolve ambiguous answers before acting on them, and keep each buyer's account distinct rather than flattening differences. Write a recommendation the team can verify. Never invent reassuring claims the business cannot back, such as a promised response time.