Data Analytics at Smarsh, referred to as Conversation Analytics, involves analyzing raw voice and multi-channel conversation data to extract meaningful, actionable insights. The Analytics page provides a dashboard that reflects voice data intelligence, integrating up to four sources: voice calls, emails, social media, and chat. This multi-channel approach accelerates insights tailored to specific business needs and allows for extensive customization of dashboards to meet unique operational requirements.
Key Metrics Categories
The following are the various metrics that you can measure:
Interactions
Interactions: Back-and-forth exchanges across audio, email, and chat. These conversations are analyzed to understand how happy or frustrated someone is, what they are trying to ask or solve, and how well the conversation is going. This helps businesses improve their customer service and make conversations smoother. Insights from interactions help improve customer service and communication.
Interactions per Workday: Total number of interactions between agents and customers during a standard 8-hour workday (Monday to Friday).
Average Interactions: The average number of interactions within the selected date range.
Highest Interactions: The agent who handled the most calls in the selected date range.
Lowest Interactions: The agent who handled the fewest calls in the selected date range.
Short Calls: Number of calls shorter than 60 seconds (configurable threshold).
Distinct Agents for Date Period: Total number of unique agents who handled calls in the selected period.
Average Agents per Workday: The average number of agents working per day within the selected date range. This excludes weekends (Saturday and Sunday).
Total Available Time (HH:MM): Total scheduled time across all agents (8 hours per day, Monday to Friday) for the selected date range.
Productive Time (HH:MM): Total time agents actively engaged in customer calls.
Unproductive Time (HH:MM): Time agents were on calls but not actively engaging with customers. Examples include being on hold, navigating systems, or waiting for information.
Idle Time (HH:MM): Time agents were not on calls. This may include time between calls, breaks, meetings, or administrative tasks.
Sentiment Analysis
Agent Sentiment: Classification of agent tone and responses as friendly, neutral, frustrated, helpful, etc.
Most Improved: Agent with the largest positive sentiment change compared to the previous month.
Most Regressed: Agent with the largest negative sentiment change compared to the previous month.
Negative Engagements: Interactions where the agent appeared frustrated, unhelpful, or disengaged.
Positive Engagements: Interactions where the agent demonstrated warmth, support, and professionalism.
Caller Sentiment: Classification of customer tone and expressions as happy, frustrated, confused, angry, or satisfied.
Call Handling Time
Productive Time: The time agents are actively engaged in conversations with customers during calls.
Unproductive Time: The periods when agents are on calls but not actively interacting with customers. Examples include:
Holding for Information: Waiting for a supervisor’s response or retrieving customer data.
Internal System Navigation: Using internal tools or software while the customer waits.
Customer Silence: Pauses when the customer is silent, either thinking or placed on hold.
Agent Performance
AQA Score: Agent’s overall quality score for the selected period, based on QA scorecards.
AQA Score MTD: Average QA score for the current month-to-date.
AQA Score MoM%: Percentage change in QA score compared to the previous month.
Best AQA: Agent with the highest QA score.
Worst AQA: Agent with the lowest QA score.
Best Question: QA scorecard question with the highest performance.
Worst Question: QA scorecard question with the lowest performance.
Best Section: QA scorecard section with the highest performance.
Worst Section: QA scorecard section with the lowest performance.
Customer Experience
Customer Confusion: Frequency of customer confusion during interactions. A higher value indicates more instances of misunderstanding.
Call Backs: Frequency of customers calling back for the same issue, indicating unresolved problems.
Profanity: Frequency of customers using offensive language, reflecting frustration or dissatisfaction.
Escalation: Frequency of calls escalated to a supervisor or higher authority.
Transfers: Frequency of calls transferred between agents or departments.
Interaction Time
Total Talk Time: Total duration of active conversations between agents and customers.
Total Hold Time: Total duration customers were placed on hold.
Total Overtalk Time: Instances where both the agent and customer spoke simultaneously.
Business Risks
Escalation: Issues requiring supervisor or specialist involvement.
Cancellations: Customers expressing intent to cancel a product, service, or order.
Formal Complaint: Customers initiating a formal complaint process.
Legal Mention: Customers referencing legal action, regulations, or rights.
Report to Agency: Customers threatening to report issues to a regulatory body or consumer protection agency.