Understanding Sentiment Analysis

Statusbrew’s Sentiment Analysis feature reads messages in context, extracts opinions and sentiments from the text, and categorizes them as positive, neutral, negative, or mixed.

Sentiment analysis in Statusbrew can help you understand customer emotions, prioritize your incoming messages and reduce noise in the Engage Inbox, track brand health, and identify potential crises.

Note: Sentiment analysis is available on the Premium and Enterprise plans only. If conversations aren't being automatically labeled as Positive, Negative, Neutral, or Mixed, check that your are on one of these plans. See full pricing details.

Use Cases Of Sentiment Analysis

• Helps you prioritize messages to engage better with your customers.

• Helps you compute satisfaction for your customers.

• Helps you create reporting dashboards to understand current sentiment toward the brand.

• Helps you monitor your campaign performance by analyzing the sentiments attached to incoming messages during the campaign.

Sentiment Categories

• Positive: Indicates a happy or favorable sentiment.

• Negative: Indicates an unhappy or unfavorable sentiment.

• Neutral: Indicates neither positive nor negative sentiment, such as a factual statement or question.

• Mixed: Indicates a sentiment that is a mix of both positive and negative.

Sentiment Score

Statusbrew’s AI examines the entire text and assigns probability scores for each sentiment label. For example, when the AI analyzes the text "I expected the product to be great but was disappointed by its performance," it might generate the following probability scores:

• Positive: 0.10 confidence

• Negative: 0.84 confidence

• Neutral: 0.05 confidence

• Mixed: 0 confidence

In this case, Statusbrew identifies the text as negative because that label has the highest probability score.

When you hover over a conversation’s sentiment icon in the conversation detail space, you will be able to find this score.

Supported Languages

The following is the list of supported languages –

How Do I Use Sentiment In The Engage Inbox?

Filtering Messages By Sentiment

You can filter messages in Engage by sentiment and create custom views based on sentiment. To filter messages by sentiment:

1. Go to Engage.

2. From the sidebar, click (Create view).

3. Click Create Manually.

4. Add an appropriate Name. You can also change the icon.

5. Click Add filter > Sentiment > In (=) > Select the sentiment you want to filter conversations for (from Positive, Negative, Neutral or Mixed)

6. Under Share access with, choose the view’s shareability from All team members, Specific user(s)/user group(s), or Only me. By default, All team members option is selected.

7. Click Save.

You can also add other filters.

Tip: In step 2, you can also select the pre-built Negative Brand Mentions view template instead of creating one manually to create a view of all brand mentions having negative sentiment. This saves time by auto-filling the required filters. You can then make edits to the view settings (if required) and save it to create the view.

Reclassifying Sentiment

While Statusbrew's AI-powered sentiment analysis is an excellent tool for gauging audience perception, the nuances of human language can sometimes misinterpret a message's intended sentiment, such as sarcastic messages.

If you come across a single message with an incorrect sentiment classification, you can refine its sentiment by following these steps, leading to more accurate insights.

1. Find the message whose sentiment you want to change and click on its current sentiment icon ( for Positive sentiment, for Negative sentiment, for Neutral sentiment, or for Mixed sentiment).

2. Now choose the correct sentiment of the message, and you will see that the sentiment icon of the message has been corrected.

FAQs

Can I create Rules based on sentiment?

Yes, sentiment is one of the filters in Rule. Read this article to learn how to create Rules.

Can I manually update a comment's sentiment?

Yes. Open the conversation in Engage, then click the sentiment icon ( / / / ) in the conversation details panel on the right. Select the correct sentiment from the four options: Positive, Negative, Neutral, or Mixed, and it updates immediately.

Can I automatically set all incoming comments to a default sentiment?

No. Statusbrew does not support setting a default sentiment for incoming comments. Sentiment is assigned automatically by the AI based on message content. If a comment is misclassified, you can correct it manually per conversation (see above).

How can I track sentiments for GIF comments?

How can I track sentiment for GIF comments?

Statusbrew's sentiment analysis reads text to determine sentiment.

GIF-only comments contain no text, so the AI defaults them to Neutral. You can still track GIF sentiment by creating a tag and manually changing its sentiment.

To track sentiment on GIF comments manually:

• Go to Engage and open any GIF comment

• Click the sentiment icon and change it to the sentiment you think is appropriate, such as Positive or Negative

• Click Tags and apply a tag like GIF to the message

Once you've tagged and re-classified your GIF comments, you can report on them:

• Go to Reports and open or create a custom report

• Click Add widget and select Create a custom widget

• Select Pie chart as the visualization

• Add Message Count as the metric and Sentiment as the segment/dimension

• Under Filters, apply the GIF tag

• Save the widget

Why doesn’t a conversation appear in the correct sentiment view after I change a comment’s sentiment?

Sentiment-based views in Engage depend on the conversation sentiment, not only the sentiment of an individual comment/message.

We keep them separate because a conversation can evolve over time. For example, a customer may start with a negative comment, but after your team replies and resolves the issue, the overall conversation sentiment may become positive or neutral.

So, if a comment is auto-detected as having the correct sentiment, the conversation sentiment usually reflects it correctly.

But if you manually change the sentiment of a comment, the conversation may still not appear in the matching sentiment-based view unless the conversation sentiment is also updated.

To fix this:

• Go to the conversation

• Open the Details panel on the right-hand side

• Update the conversation sentiment to the correct value

• Then check the relevant sentiment-based view again

For example, if you want the conversation to appear in a Negative Sentiment view, the conversation sentiment in the Details panel should be set to Negative.