You asked for feedback, and now you have too much of it. A folder of responses, a spreadsheet that keeps growing, a stack of cards from the counter. Some of it is glowing. Some of it is annoyed. Most of it sits somewhere in the middle. And you have maybe an hour before the next thing needs your attention.
So where do you start?
The honest answer is that most people start with whatever is loudest. The angriest message. The one-star review that stings. The customer who emailed three times about the same thing. That instinct feels responsible, like you're dealing with the biggest fire first. But loud and important are different measurements, and treating them as the same one will quietly steer your whole week toward the wrong problems.
Here is roughly how I'd sort a pile of feedback when time is short.
Frequency, but count it properly
The thing that comes up again and again usually matters more than the thing that comes up once at high volume. If eleven people mention slow service and one person writes three paragraphs about the font on your menu, the slow service wins, even though the font rant is more entertaining to read.
The catch is that counting is harder than it sounds. People describe the same underlying problem in completely different words. "Took forever to get seated," "no one greeted us," and "we stood at the door for ages" are three phrasings of one issue, and if you're scanning quickly they read as three separate comments about three separate things. Miss that and you'll undercount your biggest problem because it never used the same words twice.
This is the boring, grinding part of feedback work, and it's the part that decides whether the rest of your prioritisation is built on anything real. If you're doing it by hand, read for themes rather than keywords, and keep a running tally somewhere. If you'd rather not, this is exactly what AI-assisted feedback analysis is good at: grouping responses that mean the same thing even when they don't say it the same way. Qria does this across both your form responses and your public reviews, so a complaint that shows up twice on Google and four times in your own survey gets counted as six signals of one problem rather than six unrelated notes.
Once you can actually see the counts, the loudest single message often shrinks back to its real size. Sometimes it's a genuine pattern. Sometimes it's one person having a bad day, and the pile was trying to tell you that all along. There's more on what a repeated complaint is really worth if you want to sit with that idea longer.
Severity, which frequency won't tell you
Frequency tells you what's common. It says nothing about what's costing you.
A small annoyance mentioned twenty times can be less urgent than a serious problem mentioned twice, if that serious problem is the reason a customer walks out and never comes back. Wobbly tables are a nuisance. A billing error that made someone feel cheated is a different category, even if only a couple of people mention it, because the ones who felt cheated mostly didn't fill in your form at all. They just left.
So after you've counted, ask what each theme actually does to the customer. Does it mildly irritate them, or does it break the reason they came? Low-frequency, high-severity issues are the ones that hide well and hurt most. They rarely dominate the pile, so a purely count-based approach buries them under the small stuff people are happy to mention because it's safe to mention.
Who it came from
Not all feedback comes from people you're trying to serve. A first-time customer struggling with something obvious tells you more about your everyday experience than a power user complaining that an advanced feature moved. Both are real. They're not equally representative.
Be a little careful here, because this lens is easy to abuse. It's tempting to decide that anyone who criticises you "isn't your target customer" and file their feedback under noise. Sometimes that's fair. Often it's just a comfortable way to ignore things. The useful version of this question is narrow: does this person's situation match the situation most of your customers are in? If yes, weight it up. If they're genuinely an edge case, note it and move on, but write down why, so future-you can't pretend you never saw it.
Whether you can actually do anything about it
The last filter is the one people skip, and it's the most practical.
Some feedback is true, important, and completely outside what you can change this month. The rent on your second location, the supplier who keeps missing deliveries, the feature that would take a quarter of engineering time you don't have. Acting on feedback you can't act on yet is a good way to spend your hour feeling busy and finish it having changed nothing.
Separate the pile into "could fix this week," "could fix eventually," and "can't fix, but should remember." The first group is where your hour goes. The second group belongs on a list you actually revisit. The third group isn't wasted, it's context, and it changes how you read everything else. There's a fuller version of this in what to do with feedback you can't act on yet, because the "not yet" pile is where a lot of good information goes to quietly die.
Putting it together
None of these four lenses gives you a clean ranked list on its own. You hold them at once. The thing worth doing first is usually common enough to matter, severe enough to hurt, coming from people who look like your actual customers, and small enough that you can move on it before the week gets away from you. When something scores high on all four, you've found your starting point, and you didn't need it to be the loudest thing in the folder.
The pile will still be there tomorrow. It regenerates faster than you can clear it, which is a good sign, it means people are still telling you things. The goal was never to reach the bottom. The goal is to make sure that the hour you do spend lands on the responses that change how the place runs, instead of the ones that just happened to shout. If you're setting up feedback collection from scratch and want the wider picture first, the small business guide to customer feedback covers where all this fits.
