A customer satisfaction score can tell you that people struggled to use a service. The comments can tell you where they got stuck. The challenge comes when there are hundreds of responses to read, each describing the experience in slightly different terms.
Our guide to using open and closed survey questions together looks at when to invite people to answer in their own words. Here’s what to do with those answers once the survey closes.
Start with the question people answered
Keep the wording of the question alongside the comments. If people were asked what could be improved, you’d expect the answers to focus on problems. If only dissatisfied customers were shown the question, that matters too. The comments can help explain dissatisfaction among that group. To understand how common those problems were across all customers, you’d need evidence from the wider group.
Read a varied set of responses before deciding how to group them. For example, if you’re reviewing feedback on a repairs service, include comments from different repair types and satisfaction ratings. That first read can surface issues you might miss if you start with a list based entirely on what the team expected to hear. If a question consistently produces vague answers, it’s worth revisiting how the question is worded for the next survey.
Give each theme a clear definition
Use short labels that point to something specific. In a repairs survey, “appointment availability” and “updates after booking” tell a service team more than a broad label like “communication”. Write a brief definition and an example for each label so the same rule applies across responses. If two people are reviewing the comments, ask them to label the same small batch and compare their choices before splitting the rest.
One answer can cover several topics. A tenant who says, “The repair was done well, but I had to ring twice to find out when someone was coming,” is describing both the quality of the work and the lack of an update. Keep both. If you change a label halfway through, revisit earlier comments so they’re assessed by the same rule.
Count mentions and explain who answered
Say who the figures refer to. For example, if 30 out of 120 people who answered the comment question mentioned missed updates, write exactly that. If 400 people completed the survey, the result still describes the 120 who left an answer to that question. A blank comment tells you little about what someone experienced.
Count each person once per theme, even if they mention it several times. Because a response can cover more than one theme, the percentages may add up to more than 100%. Keep an eye on less common issues as well. A handful of comments about an inaccessible appointment process may warrant follow-up even if it sits low in the list.
Use quotes to explain what a theme means
Choose a short quote that illustrates a pattern in the responses. If the experiences within a theme differ, show that variation. Some people might want an earlier update while others might need an alternative to text messages. Remove identifying details before sharing comments and use a summary if someone could still be recognised from the circumstances they describe.
Read the themes alongside the survey scores and any relevant service data. If low appointment ratings coincide with repeated mentions of missed updates, check when messages were sent and what they said. That gives the team a practical way to test the explanation. Our article on the difference between data and insight explores how that extra context can change the decision you make.
Give the team something they can test
Take one clear finding and agree what to try, who will lead it and how you’ll check whether it helped. For instance, a repairs team could trial an automatic update when the appointment window changes, then review follow-up calls and survey comments for the trial period. Keep the original question and coding notes handy so the next round of feedback can be read on the same basis.
Viewpoint helps organisations design research and make sense of what customers, tenants and service users tell them. Explore our research services or get in touch to discuss your next survey.
