Most service desks are sitting on a feedback problem they don't know they have. Not a lack of feedback. Plenty of that. Tickets closed, surveys sent, responses collected.
The problem is what happens next.
A user gives you three stars. Was that mild frustration? Genuine disappointment? Sarcasm from someone who actually thought the service was fine? You don't know - because a number doesn't tell you that.
And if you don't know why the score is what it is, you can't do much with it.
Here's how feedback typically flows in a service desk:
It's not that people aren't trying. It's that the data doesn't give you enough to act on.
A 3.8 average CSAT tells you something is off. It does not tell you what is off. Is it response times? Communication quality? A specific team? A recurring issue pattern? The score hides all of that.
Meanwhile, the written comments - the part where users actually say what they think - either get skimmed once a week or pile up unread. There are too many to process consistently, and there's no reliable way to spot patterns across hundreds of freeform responses without spending hours doing it.
So the insight sits there. Uncaptured. Unused.
Think about the last time you left a review online. The star rating was almost an afterthought - the real information was in what you wrote.
The same is true in service desk feedback. Users often give a middling score but leave a comment that reveals they were actually quite happy, or vice versa - a polite four stars with a note that suggests they were frustrated but didn't want to cause a fuss.
A score without the sentiment behind it is incomplete data. You're measuring the surface, not the substance.
The result is a CSAT number that looks clean in a dashboard but masks the real picture of how users feel about your service.
When sentiment analysis isn't part of your process, a few things tend to happen:
Outliers get missed. A user who's genuinely upset might score you three stars - the same as someone who was vaguely underwhelmed. Without reading the comments, both look the same. One probably needs a follow-up.
Trends go unnoticed until they're problems. If five users in a week mentioned that resolution notes were unclear, that pattern could have been caught early. Instead, it shows up as a gradual CSAT decline with no obvious cause.
Teams act on assumptions. Without reliable insight, service improvements are often based on gut feeling or the most recent loud complaint - not what the data actually shows.
The shift needed isn't dramatic. You don't need more surveys, more responses, or a new process from scratch. You need to understand what the responses you already have are actually telling you.
Specifically: is that three-star rating mild indifference or quiet frustration? Does the four-star comment suggest a happy user or someone who didn't want to make a fuss? Is the pattern you're seeing in CSAT this month linked to a specific team, a recurring issue type, or the way updates are being communicated?
Those answers are already in your data. They just need to be read properly - consistently, across every response, not just the ones someone gets around to.
That's what Survey Sentiment, built into Solvyr® in Sunrise, is designed to do. It automatically analyses the feeling behind every survey response, so nothing slips through and you always know where you actually stand. At around £0.50 per user per year, it's one of the lowest-barrier improvements a service desk can make.
Find Out More About Survey Sentiment →