How to interpret sentiment scores
Each comment in an analysis carries a sentiment reading on a five-level scale, from Positive to Negative with Neutral in the middle. Siftsy reads sentiment three ways: as Overall sentiment, and separately as how commenters feel about the content (the post and its creator) and about the topic (the subjects being discussed).
To view sentiment scores, you need to create an analysis first. See How to start analyzing comments
The five sentiment levels
Every comment is scored on one of five levels. The three outer levels read like this:
Level | What it captures |
|---|---|
Positive | Approval, excitement, agreement, or satisfaction: praise, thanks, expressions of love or enjoyment, constructive support |
Lean positive | Sits between Neutral and Positive on the scale |
Neutral | Informational or ambiguous comments that neither praise nor criticize: questions, simple reactions, factual statements |
Lean negative | Sits between Neutral and Negative on the scale |
Negative | Criticism, disappointment, disagreement, or frustration: complaints, product issues, objections, hostile language |
If a comment's classification looks wrong, you can correct it yourself. See How to update comment sentiment.
Overall, Content, and Topic sentiment
The same five levels apply to three separate readings, so a single comment can feel different about the post than about what the post discusses:
Content sentiment: how commenters feel about this post and its creator
Topic sentiment: how commenters feel about the subjects being discussed
Overall sentiment: the two readings combined, averaged equally
The split is where the diagnosis lives. A high Content reading with a low Topic reading means the audience likes the post but not the subject: revisit the message, not the creative. The reverse means the subject resonates and the post is what's missing.
The Content and Topic readings are part of the newer scoring framework, and Siftsy Score V3 is the default. See Intro to Siftsy Score V3.
Reading the sentiment breakdown
Your analysis shows a sentiment distribution: what share of comments falls into each level. This helps you quickly assess:
Whether the audience is overall positive, negative, or mixed
How polarized the conversation is (high shares on both ends suggest division)
Whether neutral comments dominate (often a sign of informational or low-engagement content)
A balanced distribution with a slight positive lean is typical for healthy, engaged audiences. A heavy negative skew may indicate controversy, backlash, or product issues worth investigating.
Sentiment and the Siftsy Score
Sentiment is one component of the Siftsy Score, and it carries the most weight: 66%, with relevance and consensus at 16.5% each. Relevance measures how on-topic the comments are, and consensus measures how aligned commenters are with each other.
Two posts with identical sentiment distributions can have different Siftsy Scores if one has more engaged, on-topic discussion and the other has more noise or spam.
Filter comments by sentiment
Use sentiment filters to isolate one reaction and see what drives it, for example what positive praise has in common or what themes sit behind negative feedback.
Open an analysis
Open Comment filters
Under Sentiment, choose the levels you want to include
Choose whether to filter on Overall sentiment or on the Content and Topic readings
Click Apply
The comments narrow to the sentiment you selected, and the insights for that group show what the reaction has in common. For the full set of comment filters, see How to filter comments.
Combine a sentiment filter with a keyword, username, or topic search for a more targeted view. See How to explore and search the comments
What's next
Read the score and its surrounding insight cards in How to understand comment insights, or see how sentiment feeds the overall number in Intro to Siftsy Score V3.