# Understanding Sentiment

> Learn how Kommon Poll scores sentiment, subjectivity, and trends across platforms and over time.

> Learn how Kommon Poll scores sentiment, subjectivity, and trends across platforms and over time.

Numbers tell you **how much** people are talking.

Sentiment, emotion, and intent tell you **how they feel** and **why** they’re talking.

Kommon Poll automatically analyses the text of each mention to estimate:

- Whether it is positive, negative, or neutral.
- How strongly opinionated it is (subjective) vs factual (objective).
- How sentiment changes over time and across sources.

## 6.1.1 Sentiment Polarity

Sentiment polarity is the basic positive–negative–neutral classification of each mention.

- **Positive** – praise, satisfaction, recommendations, excitement.
- **Negative** – complaints, criticism, dissatisfaction, concern.
- **Neutral** – factual statements, news reports, or content with no clear opinion.

You can view sentiment at multiple levels:

- Overall distribution (for example, 65% positive, 20% neutral, 15% negative).
- Broken down by platform or source type.
- Broken down by project, country, language, or product (using filters).

Use polarity to answer questions like:

- “Is this campaign being received positively?”
- “Has sentiment improved since we responded to the crisis?”
- “Which channels are most critical or supportive?”

> Tip: Always look at the absolute counts as well as percentages. A change from 2 to 4 negative mentions is a 100% increase, but may not be serious in context.

## 6.1.2 Sentiment Subjectivity

Subjectivity measures how opinion-based a mention is.

- **High subjectivity** – personal feelings, opinions, or emotional reactions.
- **Low subjectivity** – facts, announcements, neutral descriptions.

Subjectivity is useful to:

- Distinguish opinion-heavy feedback from neutral information.
- Prioritise highly subjective negative mentions (these often drive public perception).

**Examples:**

- “Kommon Poll is the best social listening tool we’ve used.” → High subjectivity, positive.
- “Kommon Poll launched a new AI feature today.” → Low subjectivity, neutral.

## 6.1.3 Sentiment Trends by Source

Sentiment can vary significantly by platform:

- **Reviews** often skew more negative (people post when something goes wrong).
- **Instagram** might have more positive, aspirational content.
- **News sites** may be neutral but trigger emotional social reactions.

Kommon Poll allows you to:

- Compare sentiment by platform (for example, Facebook vs Instagram vs Reviews).
- Compare sentiment by source type (social vs non-social vs reviews).
- Compare sentiment by region or language.

Use this to:

- Identify where negative sentiment is strongest (and focus your effort there).
- Understand which platforms are best for testimonials vs issue resolution.
- Align channel strategies with audience sentiment patterns.

## 6.1.4 Sentiment Count & Volume Context

Beyond percentages and charts, it’s important to know how many mentions fall into each sentiment category.

- **Sentiment Count** shows the raw number of positive, negative, and neutral mentions.
- You can view trends over time to see whether volumes are climbing or declining.

Combine this with mention count and reach to answer:

- “Did sentiment get more negative because there were more negative mentions, or just fewer positive ones?”
- “Are negative mentions coming from high-reach posts, or from smaller accounts?”

</tip>