How are Review Sentiments Predicted in ProManage?

Customer reviews often contain valuable insights into business performance. To help businesses understand customer perception at scale, ProManage leverages a trained AI model powered by proprietary technology to predict the sentiment of each review. This enables businesses to quickly identify strengths, address concerns, and track improvements over time.

How Sentiment Prediction Works

AI-Powered Analysis

1. The AI model scans review text and identifies topics (e.g., Service, Staff, Cleanliness).
2. Sentiment is predicted based on keywords and phrases used by the customer in context.

Predefined Attributes

ProManage analyzes reviews across five core business dimensions (Attributes):

1. Service
2. Staff
3. Cleanliness
4. Price
5. Overall - which tells the overall sentiment.

These are the default topics most relevant to businesses. Additional custom attributes can be configured if required.

Rating-Based Prediction (No Review Text Provided)

If a customer submits only a star rating without text, sentiment is derived from the rating:

1. ⭐ 1 & 2 → Negative
2. ⭐ 3 → Neutral
3. ⭐ 4 & 5 → Positive

Text + Rating Combination

1. When a review includes both text and a star rating, the text content is prioritized for sentiment prediction.
2. The AI model analyzes wording patterns and combines them with the rating to deliver the most accurate sentiment outcome.

Sentiment Categories

Each review is categorized into one of three sentiment buckets:

1. Positive
2. Neutral
3. Negative

This helps businesses sort, filter, and act on customer feedback efficiently.

Where to View Sentiment Insights

Navigate to Insights → Reviews in ProManage.

For the selected location(s) and date range, you can view sentiment insights across:

1. Overall Sentiment (of a review)
2. Service
3. Staff
4. Cleanliness
5. Price
6. Custom Attributes (if configured)

Key Notes

1. If both star rating and text exist → text-based prediction is prioritized.
2. Default review attributes are Overall, Service, Staff, Cleanliness, and Price.
3. Custom attributes can be added on request for industry-specific sentiment analysis.

With ProManage Sentiment Prediction, businesses can move beyond raw star ratings and uncover actionable insights that drive customer satisfaction, operational improvements, and better online reputation management.