Cross-Mining
Sentiment Analysis: What Your Customers Are Feeling, Sentence by Sentence.
Cross-Mining Automatically assigns a sentiment score from 0 to 100 to every utterance in your calls, chats, and emails—from both the customer’s and the agent’s perspectives. You can identify sources of frustration, drops in satisfaction, and trends without having to listen to the recordings again.
An easy-to-read score, calculated for each sentence.
The Cross-Mining interface automatically displays a score between 0 and 100 for each utterance in your interactions. 100 corresponds to an entirely positive sentiment, and 0 to an entirely negative sentiment. The score is calculated through a classification process based on the detected words and phrases, and works for all analyzed languages.
The Customer's Experience
- Frustration, doubt, or satisfaction at each stage of the interaction
- A drop in sentiment following a key moment: the announcement of a delay, a price, or a rejection
- Priority follow-up for sensitive clients
What the counselor is going through
- Conviction, hesitation, or tension in speech
- Calls That Are Difficult to Debrief or Handle
- Consolidated scores by interaction and by advisor
Beyond positive and negative— the signals that matter.
Emotional Lows
The exact moment when the conversation takes a turn, and what was just said.
Glossary of Risky Terms
Prohibited words, aggressive language, and signs that the customer is becoming irritated.
Adjustable thresholds
Targeted thresholds and analysis windows, configured according to your business requirements.
Alerts
When a threshold is crossed, the information is automatically forwarded to the appropriate contact.
From the contact center to the entire company.
There are many use cases for sentiment analysis. Here are a few of the most common ones among our clients.
Improving Customer Service
Focus wiretapping on high-stakes calls, prepare debriefings, and monitor the impact of the actions taken.
Quality Monitoring softwareSupporting Advisors
Identify the challenges teams face in order to better train and support them.
Read the articleAmplifying the Customer's Voice
Brand experience, trends, product development, marketing, and reputation—based on what your customers are really saying.
Customer Feedback"We can sample calls and easily apply our listening checklists to them, and the Speech Analytics module performs very well."
RATP
Sentiment Analysis: Your Questions.
What is call sentiment analysis?
It is the automatic measurement of the emotional tone of a conversation. Cross-Mining assigns each utterance a score from 0 (very negative) to 100 (very positive), both from the customer’s perspective and the advisor’s, and then tracks how that score changes throughout the conversation.
How is the score calculated?
The score is generated through a classification process based on the words and phrases detected in each sentence. It is applied to calls once they have been transcribed by the speech to text, as well as to chats and emails.
Does the analysis work in multiple languages?
Yes, the feature is multilingual. To consolidate analyses conducted in multiple languages, you can also use the interaction translation feature.
Is intuition enough to judge the quality of a call?
No. It combines conversation categorization (what is said) and paraverbal cues (how it sounds) to provide a comprehensive analysis of each exchange, and then feeds into your evaluations at Quality Monitoring.
Other Speech Analytics features
Want to work with us?
Contact us
