Glossary

Sentiment analysis

Labelling comments and messages as positive, neutral or negative, so a team sees the mood and finds problems quickly.

Sentiment analysis reads incoming comments and messages and labels each one positive, neutral or negative. Over a month, it shows whether a launch was welcomed or a change upset people; day to day, it brings unhappy customers to the top of the inbox.

Machines misread sarcasm, slang and mixed messages, so treat labels as a guide for sorting, not a verdict. Look at the trend and at the negative messages themselves, and answer people rather than the score.

Put the words to work on your next week of posts

A golden honeycomb holding small photographs of social media posts