Finding Clearer Signals in User Research Interviews

User interviews are often treated as a direct route to customer truth. A participant describes what they need, why they chose a product, and what they might do next. Yet every answer is shaped by memory, social expectations, the wording of the question, and the story the person has already formed about their own behaviour.

Confirmation bias makes this especially tricky. Researchers may unconsciously favour evidence that supports an existing hypothesis, while participants may selectively recall experiences that fit their beliefs. In Australian service research, where a relaxed “no worries” can mask genuine friction, careful interview design matters.

How Confirmation Bias Enters The Room

Confirmation bias is the tendency to notice, interpret, and remember information in ways that support an existing view. In qualitative research, it can appear before the first interview, when a team writes a discussion guide around the assumption that a feature is valuable or a service problem has a single cause.

It can also emerge during the conversation. A researcher might probe deeply when a participant agrees with the hypothesis, then move quickly past a contradictory answer. A customer who says an online claims process is “pretty easy” may receive a polite follow-up, while someone who complains may be asked for detailed examples. The result is an uneven evidence base.

Australian communication styles can add another layer. Participants may soften criticism with phrases such as “a bit average” or “not ideal”, particularly when speaking to a researcher from a respected organisation. Taking language literally can cause a serious usability problem to be coded as a minor inconvenience.

Separate Discovery From Proof

A strong interview is designed to discover how people behave, rather than to prove that a solution is sensible. Replace leading questions such as “Would faster delivery make you use us more?” with prompts about the last relevant occasion: “Tell me about the last time you needed this service. What happened?”

Past behaviour usually provides more reliable insight than predictions. Ask what the participant did, what alternatives they considered, who influenced the decision, and what happened afterwards. If someone says they would “definitely recommend” a service, explore the circumstances in which they have actually recommended a business.

Social influence deserves particular care. Participants may claim that trust comes from professional credentials, while their real choice is shaped by reviews or recommendations. Research on social proof and trust can help teams examine this gap without treating stated preferences as final evidence.

Build A Discussion Guide That Challenges You

A balanced guide includes questions that could disconfirm the team’s working theory. If the hypothesis is that customers abandon a form because it is too long, ask what else was happening at the time: poor mobile reception, missing documents, uncertainty about eligibility, or a competing task.

Use neutral wording and a consistent sequence across interviews. Begin with context, move to a recent experience, then explore interpretation and emotion. Avoid introducing the proposed solution too early, since showing a prototype can make participants search for ways to be helpful.

Useful safeguards include:

• Write the hypothesis and its possible alternatives before recruitment.
• Ask for concrete examples before opinions or forecasts.
• Include participants who do not fit the expected customer profile.
• Record exact phrases rather than translating them into assumptions.
• Mark uncertainty clearly during analysis.

This approach is relevant to Australian organisations serving very different contexts, from a Sydney fintech customer on a mobile device to a regional business working with patchy connectivity. A single urban user journey should not quietly become the national customer experience.

Analyse Evidence Without Chasing Agreement

Confirmation bias often becomes strongest after the interviews, when teams cluster notes on a wall or review transcripts at speed. A memorable quote can dominate the discussion, especially when it supports a senior stakeholder’s preferred direction. Frequency helps, but it is not the only measure of importance.

Separate observations, interpretations, and recommendations. “The participant abandoned the application after the identity check” is an observation. “Identity verification is the main barrier” is an interpretation that needs comparison with other cases. “Remove the identity check” is a recommendation that requires feasibility, risk, and regulatory review.

A structured coding process can include disconfirming evidence as a deliberate category. Ask which participants behaved differently, what evidence would weaken the current explanation, and whether a finding applies to a segment or to the broader market. In a country with distinct metropolitan, regional, and remote conditions, averages can conceal meaningful differences.

Turn Interview Learning Into Better Decisions

Research findings become useful when they change a decision, not when they simply produce an attractive report. For each insight, specify the customer behaviour involved, the evidence supporting it, the confidence level, and the next test. This prevents a persuasive anecdote from being treated as a universal rule.

Workshops can help cross-functional teams examine competing interpretations, provided participants retain the important details. Guidance on making workshops memorable is valuable when research findings need to travel from a project team to frontline staff, product managers, and executives.

A practical evidence review can ask:

• What did people actually do, rather than say they might do?
• Which findings appeared across different customer groups?
• What evidence contradicted our preferred explanation?
• What would we expect to observe if the insight were wrong?
• Which small service change can be tested safely?

For start-ups and established organisations alike, external perspectives can expose blind spots. A programme such as DHL Startup Lab reflects the value of testing ideas against real operational and customer conditions rather than relying on internal enthusiasm.

Good user research does not remove judgement; it disciplines it. By treating interviews as evidence to interpret, challenge, and triangulate, Australian teams can design services that reflect real behaviour across city, suburban, regional, and remote settings. Engage an applied behavioural science researcher or equip your team with a bias-aware interview process before the next round of customer conversations.