Sentiment
in Customer and Agent Interactions
How Does Sentiment Work?
How Is Sentiment Used?
What Is Sentiment?
PART 1 OF 2
Sentiment is a machine learning (AI) model trained with a large, comprehensive CX dataset to measure whether a customer interaction is positive, negative, or neutral on a relational scale.
It can be used to:
understand call volume and trends
measure agent performance
identify process or product issues
provide live guidance to agents during conversations with customers
Phone calls
Emails
Chats
Social media posts
Any voice or text interaction can be analyzed in real-time by AI modules for insights into where customer sentiment is low or high—and why.
Sentiment also identifies interactions that start negative and turn positive, or the other way around.
-
+
1
2
Semantics, because positive phrases can offset negative ones
Context, because some words that may normally be used positively can also indicate sarcasm and frustration
The model analyzes other factors, too.
These additions improve accuracy because they give the model more knowledge to work with.
Changes in pitch, tone, or speaking rate, which can signal change in satisfaction
Cross talk, which could indicate confusion or frustration
Laughter, which could indicate a positive change
The model identifies positive and negative words and phrases, whether spoken or text, but it considers two key factors:
Language is key.
Understanding volume and trends is the real value behind sentiment analysis, and there are many use cases where it can deliver value.
Customer Satisfaction
3 Key Elements to Put Customer Satisfaction in Action
VIDEO See how Enlighten AI creates CSAT in an out-of-the-box solution
BROCHURE Enlighten AI for Customer Satisfaction
Sentiment scoring is a proven predictive indicator of customer satisfaction such as NPS, tNPS, and CSAT surveys. But at a deeper level, it is the agent behavior during an interaction that can positively or negatively influence the sentiment score on every interaction they have with a customer. Enlighten AI for Customer Satisfaction is an out-of-the-box solution that objectively measures agent behaviors proven to drive customer satisfaction consistently and accurately, on every interaction, with no bias. This provides agents and managers with metrics they can trust.
For example, the Sentiment score above begins at 50, a neutral score, and ends at 100, a positive score. John used the prompts during his conversation to ensure a positive experience for his customer.
VIDEO Watch a demo of how it all works.
BROCHURE Enlighten AI for Customer Satisfaction with Real-Time Interaction Guidance
Customer satisfaction behavioral models in real time.
Real Time
Scoring how specific agent behaviors affect customer sentiment is truly revolutionary. But for the first time ever, businesses are driving customer satisfaction on every single interaction, in real time. Real Time Interaction Guidance accurately and immediately scores the specific agent behaviors proven to drive customer satisfaction and sentiment scores. Easy-to-understand recommendations empower agents to deliver excellence, one conversation at a time. Not only is this proven to drive customer satisfaction scores, it also boosts employee satisfaction and retention through improved engagement and better understanding of how to self-improve or request personalized coaching.
Personalized Survey Feedback
Process Improvement Analysis
Sales Effectiveness
Product Performance Monitoring
Agent Enablement and Coaching
Interaction Analytics Auto-detection of words and phrases using AI on 100% of interactions
AI Predictive and Interpretive Models Auto-scoring of agent soft-skill behaviors proven to drive customer satisfaction
AI Predictive and Interprdtive Models Auto-scoring of agent soft-skill behaviors proven to drive customer satisfaction
Next-Generation Quality Management Solution
Automated workflows and evaluations
Operationalize results
Targeted coaching and dashboards
Sentiment and the NICE Enlighten AI for Customer Satisfaction behavioral models are useful for giving managers the ability to see where each team member ranks at a glance. Rank and comparisons can easily be made among individuals in a team, across teams, or even across sites.
The benefits of sentiment-based agent enablement include:
Empowers agents to self-improve on their own.
Empowers managers to deliver personalized, targeted coaching.
Focuses all roles in the organization on driving customer satisfaction.
WHITEPAPER Strategies for Advancing Your CX Program with AI-Enabled, Analytics-Driven Quality Management
Improve Team Performance
Best Practices
Coaching Opportunities
Great examples of interactions to use for coaching.
Identify behaviors to improve with coaching.
Sentiment scoring can be used from any point of view, such as by product. Use cases for searching sentiment scores by product include but are not limited to:
Collecting feedback on new products
Measuring sentiment over the product lifetime as it matures
Measuring satisfaction or dissatisfaction after a product change
Discovering product defects or ineffective customer communication
Sentiment around sales attempts can inform an organization how well sales initiatives are working and provide insight around what employees are struggling with and where they are successful. Sentiment around the sales themselves can indicate whether the initiative itself needs to be refined.
CASE STUDY
TTEC: Customer Experience Technology Services for Global Brands
— Barbara Wingle, VP of Insights and Analytics
We were able to quickly and efficiently understand VoC and VoA to make data-driven decisions that impact business performance in a positive way —and accurately measure results.
Impact
increase in agent revenue
159
increase in sales production during agent nesting
360
increase in average first-week sales production
833
NICE Solutions
NICE Engage
Desktop Analytics
Desktop Automation
Nexidia Analytics
49,700 employees
More than 40,000 agents
Locations on 6 continents
%
Just as agent behaviors can drive sentiment, so can broken processes, especially if the root cause of the problem is not clear. Measuring the right things and then examining the data can uncover a process problem. In one recent case, a service provider used sentiment to identify the most common key words in customer/agent interactions that were associated with low CSAT scores. They found a large volume of conversations were around network and wifi set up and were able to quantify the scope of the problem.
21
Result: AHT dropped by
As a result of these findings, the service provider improved self-help options for resetting wifi connections and updated agent troubleshooting steps to include reset of wifi up front.
2X Average Handling Time (AHT)
20.4
minutes
Wifi-related calls
9.6
All calls
vs.
20
Agent contacts for wifi problems
Targeted feedback based on analytics findings Complete data interpretation with data from direct, indirect, and inferred sources Informed action to close the loop with customers
Using Sentiment with Targeted Surveys
Repeat Calls
Dissatisfaction
Adaptive survey collection can use sentiment information to send targeted feedback after key interactions, such as possible churn or cancellation risks.
Sales
Using sentiment and AI in this way opens new paths to delivering improved customer experiences by providing:
Survey data tends to fall in the extreme positive or negative
The greater volume of interactions have no survey data captured
Difficult to identify trends or business value drivers
The Challenge with Traditional Survey Programs
Traditional survey collection is known to result in a dichotomy of very positive or very negative feedback, with very little data in between. Additionally, the data set is indicative of only a very small number of customer experiences. Sentiment scoring is more powerful because it scores every transaction. This helps Voice of the Customer (VOC) programs because they can adapt their survey system to send a brief survey after a negative interaction to ask the customer if they would like to be contacted by a manager to resolve their issue. Targeted, contextual surveys have a higher response rate, and can help prevent customer churn or cancellations.
Take Your Next Step.
Deliver Extraordinary CX.
When it comes to uncovering valuable insights about your customers, Enlighten AI for CX does the work for you. Enlighten AI for CX is the only one-vendor, single AI solution for accelerating action to transform entire organizations into customer-centric businesses, right out-of-the-box. Immediate results offer prescriptive insights throughout the customer lifecycle and are leveraged to optimize every moment of the customer’s journey with the business. We are the only ones who can offer an all-in-one view for effective change across every role in the organization. For the first time, all stakeholders can have the intelligence required to accelerate actions proven to improve customer satisfaction and lifetime value.
To learn more about how AI is revolutionizing customer experience and contact center analytics, visit www.nice.com/enlighten-ai.
VIEW PART 2