Sentiment score
A sentiment score is one number for how positively AI answers talk about a brand, usually its share of positive mentions.
A sentiment score is a single number that sums up how positively AI answers talk about a brand. Tools that offer one usually take the share of mentions that are positive, sometimes averaged across platforms, sometimes weighted so the things said most often count most.
It's popular because it fits on a slide. One number, up or down, easy to report each month.
Check how it's worked out
Two tools can give the same brand very different scores from the same answers. Some count neutral mentions in the total and some leave them out. Some weight each theme by how often it comes up. Some average platforms equally, however many answers each gave. Aggregated sentiment score has a worked example of how far apart the results can be.
So before you compare scores, from two tools or from two agencies, find out how each one is made.
What to read alongside it
A score tells you the overall mood. It can't tell you why, or what to do. For that you need the things being said.
For example, Kestrel Roasters' score holds steady at 80% for two months. Underneath, praise for its espresso has grown, and complaints about slow delivery have grown too. The two cancel out in the score. Only the details show that delivery has become a problem.
Read the score as a signal to look, then read the negative answers and the pages behind them.
In AnswerOrbit
We don't make a sentiment score. Every mention is marked positive, neutral or negative, and the Sentiment tab in Perception shows how many of each there are, by brand or by platform, day by day. What's said lists each thing the answers say about you, with its tone, how often it's said and the pages behind it.