Coding Open-Ended Answers
Free-text answers and complaint records cannot be counted until they are grouped. Traditionally somebody reads a few hundred of them and sorts them into a short list of categories — a code frame. SenSight asks an AI model to do that pass for you and writes the result back as an ordinary column, so Pareto and the cross-tab can use it immediately.
The important part is not the AI, it is the frame. Leave it empty and the model proposes one; paste an existing frame and the model may only assign to it — it cannot invent new categories, and anything that fits nowhere goes to "Other". That restriction is what lets you code a second wave of the same survey and still compare the two reports. A frame that drifts between runs turns a trend into noise.
Two cautions. The answers you select leave your browser and go to your AI provider, and open-ended answers often contain names and phone numbers — clean them first. And always read the examples column: if a category is a grab-bag of unrelated comments, the frame is wrong, not the data.
The important part is not the AI, it is the frame. Leave it empty and the model proposes one; paste an existing frame and the model may only assign to it — it cannot invent new categories, and anything that fits nowhere goes to "Other". That restriction is what lets you code a second wave of the same survey and still compare the two reports. A frame that drifts between runs turns a trend into noise.
Two cautions. The answers you select leave your browser and go to your AI provider, and open-ended answers often contain names and phone numbers — clean them first. And always read the examples column: if a category is a grab-bag of unrelated comments, the frame is wrong, not the data.
Try it in the app
Try: Load "Employee Survey — Engagement" → Cross-Tab tab → Open-Ended Coding → leave the frame empty → Code Answers → Add as column. Then set the cross-tab rows to the new Comment_coded column and columns to Department.