NPS Response Analyzer
NPS = (promoters − detractors) ÷ all responses × 100, rounded to one decimal. Mean score uses two decimals. User-selected keywords match literal substrings, once per feedback per keyword; keywords may overlap and counts must not be added together. Missing feedback is excluded from keyword percentages. Case-insensitive mode uses simple lowercase conversion. No sentiment, topic identification, significance, representativeness or causal inference. Inputs remain in your browser.
A QUICK WALKTHROUGH
How to use this tool
- Enter one 0–10 score per line, optionally followed by a Tab and feedback.
- Optionally choose literal keywords and case sensitivity.
- Review descriptive counts, average, NPS and keyword matches.
Descriptive results
NPS = (promoters − detractors) ÷ all responses × 100, rounded to one decimal. Mean score uses two decimals. User-selected keywords match literal substrings, once per feedback per keyword; keywords may overlap and counts must not be added together. Missing feedback is excluded from keyword percentages. Case-insensitive mode uses simple lowercase conversion. No sentiment, topic identification, significance, representativeness or causal inference. Inputs remain in your browser.
Responses: one 0–10 integer per line; optional feedback after a Tab
Limits: 200,000 input characters, 5,000 nonempty rows, 500 feedback characters per row; 20 keyword lines of 60 characters, 2,000 keyword-input characters. No headers or CSV. Blank lines are ignored. Each nonempty row counts as one response; duplicates are not removed.
GOOD TO KNOW
Common questions
How is NPS calculated?
Subtract detractors from promoters, divide by all responses, and multiply by 100. The displayed NPS is rounded to one decimal.
Can I add keyword counts together?
No. One response can match several keywords, so counts may overlap. Each keyword counts at most once per feedback entry.
Does the analysis infer sentiment or statistical significance?
No. It reports descriptive counts and literal keyword matches; it does not identify sentiment or establish significance, representativeness, or causation.