Sentiment analysis leverages AI-powered prediction to assess the emotional tone and attitude expressed in customer interactions. This process offers insights into customer sentiments, enabling real-time monitoring, trend analysis, and timely interventions to improve customer experiences, prevent churn. Sentiment analysis facilitates real-time monitoring of customer experiences, enhancing conversation prioritization, and providing comprehensive retrospective insights for support teams to improve processes and products based on customer sentiment.
In sentiment analysis, the expressed sentiments are indicated by displaying the starting, current and ending sentiments, which can be categorized as positive, negative, or neutral.
Benefits of Sentiment Analysis
Issue Prioritization: Sentiment analysis can prioritize conversations based on sentiment, allowing support teams to address urgent issues first and allocate resources effectively.
Proactive Issue Resolution: Sentiment analysis enables companies to identify and resolve customer issues promptly, reducing the likelihood of problems escalating and resulting in more positive customer experiences.
Enhanced Customer Satisfaction: By understanding and addressing customer sentiments in real-time, businesses can improve their support, leading to higher customer satisfaction levels.
Customer Churn Reduction: Real-time monitoring of sentiment helps businesses identify at-risk customers and take proactive measures to retain them, reducing customer churn.
HOW TO ENABLE OR DISABLE THE FEATURE FROM THE ADMIN PAGE:
Go to the admin settings and locate the Chat settings.
Scroll down to find the Freddy icon.
Click on the Freddy icon to open the list of AI features available.
Choose the specific feature you want to enable or disable.
Toggle the switch to the right to enable the feature or to the left to disable it.
How to use Sentiment Analysis?
Sentiments in the agent Inbox - Once you've logged in and accessed the agent inbox, sentiments for all conversations will be visible in the left-hand pane. Hovering over a sentiment will reveal the specific sentiment value assigned to each particular conversation.
After opening a particular conversation from the left pane, you can also observe the sentiment value displayed in the middle pane alongside the title of the conversation.
Step 2: Prioritizing Conversations Based on Sentiment
In the agent inbox, you can prioritize the conversations based on Sentiment. Agents have the option to filter conversation by sentiment, such as positive, neutral, or negative. This feature helps agent to focus on and address conversations that demand urgent attention.
Take into account the customer sentiment in conversations, even when categorized as medium or low priority. If the ongoing customer sentiment is negative, prioritize addressing the customer's sentiment state. Conversely, if a high-priority conversation currently reflects positive sentiment, consider directing attention to other high-priority conversations with more pressing needs.
Step 3: Address High Negative Sentiment and High Priority: Identify conversations with both high negative sentiment and high priority. Directly access the issue related to such conversations to address the customer's query and concerns.
Step 4: Provide Updates and Monitor Sentiment Changes
After providing an update or resolution, observe how the customer reacts.
If the customer acknowledges the response in a way that suggests a change in sentiment, it may transition from negative to neutral or even positive.
By following these steps, you can effectively utilize Sentiment Analysis to prioritize and address customer sentiment in your support interactions, leading to improved customer experiences and issue resolution.
Sentiment Analysis Report
For admins - The goal of the admin is to view sentiment trends to obtain insights to improve support processes/product. Create automations to prioritize conversations coming in with negative sentiments.
The sentiment analytics report provides admins with valuable insights into various aspects of customer sentiment throughout a specific time period. It enables admins to gauge the average sentiment at the beginning of customer-agent interactions, referred to as the "beginning sentiment." This initial sentiment is captured as soon as a conversation is started, and the customer adds their first reply.
The report calculates the "average sentiment for resolved conversations," which reflects the sentiment at the end of a conversation when it's resolved. Ideally, a conversation may start with a neutral or negative sentiment but should conclude with a positive sentiment, indicating a positive support experience.
By analyzing this data, admins can gain a high-level overview of sentiment trends over time. For example, if the average sentiment for resolved conversations is trending negatively, admins can identify whether this trend persists over a week or if it's changing. This comparison helps them understand how different sentiment categories are evolving based on these trends enabling them to identify and take action to address underlying issues.
Conversely, if there's a scenario where only the negative sentiment is increasing while the positive sentiment is declining, it prompts further investigation. One might delve deeper to examine the specific types of conversations involved or whether they are linked to a particular agent.
In the case of agent analysis, it becomes essential to determine how many agents successfully concluded conversations on a positive sentiment note, which essentially represents resolution sentiment. For instance, one might observe that a particular agent predominantly resolves his conversations with a negative sentiment.
To dive deeper into this, admins can examine the specific conversations that contribute to the trend. They can check for recurring themes or patterns in these conversations, especially in terms of their subject matter. This analysis can help identify any areas agents might need more support or better understanding of the processes.
Admins can also expand their view by using filters based on agent groups. For instance, agents can focus on the building group and check how the average sentiment of their conversations has changed over time. This approach also allows admins to assess the performance of all agents in that group.These insights are called "sentiment metrics”. In the rows shown in the chart, you'll see the beginning sentiment, while in the columns, you'll find the ending sentiment. For example, there are 133 tickets that both started and ended positively, while 60 conversations began positively but ended negatively. This detailed analysis offers valuable information for informed decision-making and improvements.
Admins have the capability to establish workflows based on the conversation sentiment. For instance, if the beginning sentiment equals negative sentiment, they can automatically assign it to highly skilled agents.
Alternatively, they can create workflows to trigger notifications, when sentiment shifts from positive to negative. These automations offer the flexibility to swiftly respond to changing sentiment dynamics and, if necessary, alter issue priorities to address urgent matters effectively.
*The enablement of optional functionality is subject to certain feature-specific terms and conditions set forth in the Freshworks Supplemental Terms.