Look for emotion words in a message, review or post and see how joy, sadness, anger, fear, surprise and trust compare. It works from a word list and sets aside words that follow a negation, so it's a rough guide that can't read sarcasm or context.
How to use it
- Paste your text.
- Read the breakdown.
- Check the main feeling.
- Remember it's a simple word-based estimate.
When it helps
- Getting a quick read on customer feedback.
- Checking the tone of a draft message.
- Teaching how text analysis works.
- Comparing the tone of different texts.
Good to know
- Longer text gives a steadier picture.
- Sarcasm and slang are missed.
- It only supports English.
How the counting works
The tool keeps lists of words associated with each emotion, such as delighted and grateful for joy, or terrified and panic for fear. It scans your text, counts matches for each emotion, and reports each emotion's share of the total. If one of the two words before a match is a negation, such as not or never, that match is set aside and reported separately.
This is the simplest form of sentiment analysis, known as a lexicon approach. It's transparent, since you can see which words were counted, but it's blunt, since it knows nothing about phrases or tone.
Worked example: a product review
A review saying "I was so excited to open the package, but it arrived broken and nobody answered my emails. I'm furious, though I still trust the brand to make it right" shows joy, anger and trust. That mixed result is accurate: the writer is angry but hasn't given up on the company.
Compare two versions of an apology email, one that says "we're sorry for the terrible mistake" and one that says "we apologise for the inconvenience". The first registers sadness; the second registers almost nothing, which tells you something about how warm it sounds.
Where it goes wrong
Sarcasm is the classic failure. "Great, another delay" is counted as joy. Slang, new words and words with several meanings are also missed, and there's no list for languages other than English.
Use it for a quick sense of tone, or to teach how text analysis works. Don't use it to judge people or make decisions about them.
Emotions in customer writing
Support teams often scan feedback for anger or fear, which signal urgent problems. Marketing teams look for joy and trust, which show what customers value. A quick tool like this can sort a pile of comments before anyone reads in depth.
Always read the highest-scoring messages yourself. A word list finds the vocabulary, but people supply the meaning.
Reading the percentages
The percentages are shares of the emotion words found, not measures of how strongly someone feels. A text with two anger words and no others is 100 percent anger, even though two words isn't much to go on.
Questions
How does it decide what counts?
It counts words from a list of emotion words, so it only sees the vocabulary, not the context.
Why was a word ignored?
It followed a negation such as not.