Survey Free-Text Categorizer — Free Keyword Bucket Sorter

This tool sorts open-ended survey answers into categories that you define with keywords. You write one line per category — a name followed by the words that signal it — and paste the responses, one per line. The tool assigns each response to every bucket whose keywords appear in it, and counts how many responses landed in each bucket. A response with no keyword hit lands in Uncategorized, which is itself useful information. This is keyword matching, not artificial intelligence: the tool does not understand meaning, irony, or negation. If you write the buckets well, it is a fast first pass over hundreds of answers. If you write them badly, it is a fast way to be wrong at scale.

How to use it

Write your buckets, one per line, as Name: keyword1, keyword2.
Paste the open responses, one per line.
Click Categorize and read the counts table first.
Open the Uncategorized rows: they show which keywords you forgot.
Add the missing keywords, re-run, and repeat until Uncategorized is small.

Worked example

Thirty students answered “What could be better about the course?” With buckets Price, Quality, and Service, the first run leaves twelve answers Uncategorized. Reading them shows repeated words like “schedule” and “timetable” that no bucket covers. Adding a Schedule bucket with those keywords and re-running drops Uncategorized to three answers, which turn out to be jokes and blank-ish replies. The counts now say something real: Service complaints outnumber Price complaints two to one. That insight came from reading the misses, not from the first automatic run.

Honest limits

Keyword matching is literal. “Not expensive” matches the Price bucket even though the respondent is praising value, and “the staff were slow but kind” matches Service without telling you whether the sentiment is good or bad. Sarcasm, misspellings, and synonyms you did not list all escape. Do not report keyword counts as if a human coded them. For a thesis or publication, use this tool to draft categories, then hand-code a sample of at least fifty answers and check agreement before trusting the counts.

Who it is for

This tool is for survey researchers, teachers reading course feedback, and small teams triaging open comments without a statistics package. It shines on short answers — one or two sentences — where a keyword usually carries the point. For long essays or interviews, use proper qualitative coding instead, where a human reads the whole passage in context.

Going further

A keyword pass becomes trustworthy when you build a small codebook and check it. Write one sentence per category saying what belongs and what does not, with two example responses each, and keep that codebook with your analysis. Then hand-code a random sample of fifty to one hundred answers without looking at the tool output, and compare: the share of agreements is your first quality number. Where human and tool disagree, read the response and fix the bucket — usually a missing synonym, a negation, or a category that is really two categories sharing a word. Repeat until agreement is high and stable. Report the final codebook, the keyword lists, and the agreement figure in your methods or appendix, because keyword counts without a codebook cannot be reproduced by anyone else. Also watch for category drift: as you add keywords to catch misses, a bucket can quietly grow to swallow responses that belong elsewhere. Re-check the early buckets after every round of additions. One more use of the Uncategorized pile: it is an early warning for topics you did not expect. If a new theme keeps appearing in the misses — a product fault, a scheduling complaint, a request nobody anticipated — that theme may be the most important finding in the survey, precisely because no question asked about it. Promote recurring misses into categories, but also flag them in your report as emergent themes: findings the respondents raised themselves carry different weight from answers to prompted questions, and honest reports mark that distinction instead of folding emergent themes silently into the counts.

Embed this tool on your website — free

Copy the code below to add this working tool to your own page. A small credit link stays with it — that’s all we ask.