Start with a finding, not a recommendation
Research often produces a long list of observations: a confusing question in a survey, a recurring complaint in interviews, or a difference between audience segments. The challenge is not simply to rank the most interesting items. It is to decide which findings deserve attention now, which need a closer look, and which can wait.
Write each finding as a specific statement about what the research showed. For example, “Several interview participants could not find delivery information before checkout” is easier to assess than “Improve the website.” Record where it came from, which audience it concerns, and any important limits. Keep the proposed response separate. This makes it possible to agree that a problem may exist without assuming that the first suggested solution is the right one.
Define impact in terms of decisions
Impact is the difference a finding could make if the team responded well to it. Ask which customer need or business decision it affects, how many relevant people might encounter the issue, and what the consequences may be. A small point of friction in a frequent task may matter more than a dramatic complaint about a rare situation. Conversely, an issue affecting few people may still deserve attention if the potential harm or difficulty is substantial.
Use a simple scale, such as high, medium, or low, but define it for the decision at hand. High impact might mean a finding could change a major product choice or remove a serious obstacle in a common journey. Low impact might mean it suggests a minor wording improvement. Note your reasoning beside the rating. A clear explanation is more useful than a score that appears precise but hides assumptions.
Consider whose experience is represented. An average can obscure a problem concentrated in one audience segment, while a vivid account from one participant may not reflect a widespread issue. Neither should be dismissed automatically. The question is whether the finding could change a decision the team is able to make.
Assess evidence strength separately
Evidence strength describes how much confidence the available research supports. Look at the fit between the method and the claim. Interviews can reveal how people describe a problem and why it may occur, but a few interviews cannot establish how common it is. A survey can help estimate the distribution of responses among the people reached, but a poorly worded question or an unrepresentative sample can weaken that conclusion.
Check whether the finding appeared consistently, whether different sources point in the same direction, and whether a plausible alternative explanation remains. Did participants complete a task, or only say what they expected to do? Did a segment difference persist when you looked at relevant context? Document limitations plainly, including small groups, missing audiences, ambiguous questions, and conditions unlike the real setting.
An evidence rating can also be high, medium, or low. High does not mean certain; it means the available material is strong enough for the intended decision. A finding may be well supported as a description of the participants studied but weak as a claim about all customers. Phrase the conclusion at the level the evidence permits.
Use the two ratings to choose a next step
Place findings in an impact-and-evidence grid. High-impact findings with strong evidence are good candidates for action planning. High-impact findings with weak evidence deserve prompt validation, especially if acting on a mistaken conclusion would be costly. Low-impact findings with strong evidence may fit into routine improvements. Low-impact findings with weak evidence can usually remain on a watch list rather than consume immediate effort.
The grid is a discussion aid, not an automatic decision rule. Consider the cost and reversibility of a response. A small, easily reversed change may be reasonable on modest evidence; a major redesign calls for more confidence. Also distinguish between acting on a finding and testing a proposed solution. Strong evidence that people struggle with a step does not, by itself, prove that a particular new design will help.
Make the priority list usable
For each priority finding, create a short record: the observation, the supporting evidence, the impact and evidence ratings, key uncertainties, a proposed next step, and a person or team responsible for considering it. The next step might be a design test, a follow-up interview, a survey question to refine, or a decision to make a limited change and check the result.
Review priorities when the decision changes or new evidence arrives. A finding that was minor for one project may become important for another; a tentative pattern may strengthen or disappear with further work. By keeping impact and evidence separate, teams can give urgent questions appropriate attention without presenting uncertainty as certainty—and can turn research into decisions that are easier to explain and revisit.
