State the assumption and the finding separately
Suppose your team expects customers to choose a product mainly because it saves time. In interviews, however, several customers talk first about feeling confident that they are making the right choice. The instinct may be to defend the original idea or replace it immediately with a new one. Neither response makes full use of the research.
Start by writing two plain statements: what you expected and what you observed. Then note how the finding was produced. Did participants mention confidence without prompting, or did they select it from a list of answer options? Was the question about their last purchase or about what they might buy in the future? These distinctions help you avoid treating an interpretation as if it were a direct observation.
Check the strength and scope of the evidence
A contradiction is worth investigating, but first ask how much weight the evidence can carry. Review who took part, how they were recruited, what they were asked, and whether the setting may have shaped their answers. A handful of interviews can reveal language and possible motivations; they cannot, on their own, show how common a motivation is across your audience. A survey can show patterns among its respondents, but its usefulness depends in part on question wording and who answered.
Look for variation rather than rushing toward a single conclusion. Did every participant express the same view, or was the surprising answer concentrated among people with a particular experience? Check whether the original assumption holds in some situations but not others. For example, time savings might matter when customers compare options, while confidence matters when they commit to a purchase. Both findings could be useful if they describe different moments in the decision.
Explore more than one explanation
Write down at least two plausible explanations for the result. In the product example, customers may genuinely value confidence more than speed. Alternatively, they may see time savings as a basic requirement and mention confidence because it distinguishes one option from another. A third possibility is that the interview prompts made confidence easier to discuss. Keeping these explanations separate makes the next research step clearer.
Return to the raw material where possible: interview notes, recordings, open-ended responses, or the exact survey questions. Look for what people described doing, not only the labels they used. If a participant says they want confidence, what helps them feel it: a clear comparison, an explanation of limitations, a familiar feature, or an opportunity to ask questions? Specific behaviors and situations are usually more actionable than a broad theme on its own.
Choose a follow-up that addresses the uncertainty
You do not need to repeat an entire study to learn more. Choose a follow-up based on the decision at stake. If you are considering a change to a product page, you might ask a fresh group of participants to compare versions that emphasize different benefits. If you need to understand the range of motivations, you might revise a survey so respondents can rate several reasons rather than select only one. If the finding seems limited to a particular audience, recruit people outside that group before applying it more broadly.
Frame the follow-up as a question that could change your mind. Instead of asking, “Can we confirm that confidence is the real driver?” ask, “When do people prioritize confidence over speed, and what do they mean by confidence?” This wording leaves room for a mixed answer. Record your proposed interpretation before collecting more feedback so you can recognize when the new evidence does not support it.
Turn the surprise into a proportionate decision
Research does not have to settle every uncertainty before it becomes useful. Match the action to the strength of the evidence and the cost of being wrong. A surprising interview theme might justify trying clearer reassurance in a draft message. It may not justify replacing an entire positioning strategy. A consistent pattern across several relevant sources could support a larger change, while still leaving room to monitor how people respond.
When sharing the finding, distinguish observation, interpretation, and recommendation. For example: “Participants repeatedly asked how to judge whether this option was right for them. That suggests confidence may play a larger role than we assumed. We recommend testing clearer decision guidance before changing the main product message.” This approach neither dismisses an uncomfortable result nor overstates it. It makes the contradiction a starting point for a better decision.
