Making sense of customer comments with AI, carefully
Making sense of customer comments, with care
Customer comments pile up fast. These pages explain, one idea at a time, what AI tools can do with survey answers and where a person still needs to look. They cover sorting, checking, counting and privacy.
Written for owners and staff of small businesses who have a pile of feedback and want to use it carefully.
All the lessons
What AI can and can't do with customer feedbackWhat an AI tool does with a pile of survey comments, where it goes wrong, and which jobs still need a person to read and decide.
Sorting comments into themes with an AI toolHow to turn hundreds of free-text comments into a short list of themes, with a person checking the groups before anyone relies on them.
Open-text survey questions and how AI reads themWhich survey questions give answers an AI tool can sort well, and how to word an open question so the replies are easier to group.
How to check an AI summary of customer feedbackSimple checks to confirm an AI summary matches what customers wrote, from spot-reading comments to hunting for invented quotes.
Privacy when customer feedback goes into an AI toolWhat to check before pasting customer comments with names and details into an AI tool, including what an Australian regulator advises.
Counting themes without fooling yourselfWhy an AI tool's counts of how many people said something need checking, and a plain way to count themes and read the results fairly.