Understanding Call Analytics
Read dashboard stats, understand call statuses, confidence scores, sentiment analysis, and export reports.
3 min readGetting Started
Reading Your Dashboard Stats
Your dashboard overview shows four key numbers:
| Metric | What It Means |
|---|---|
| Total Calls | Number of calls your AI handled in the selected period |
| Avg. Duration | Average length of all calls |
| Resolution Rate | Percentage of calls completed without transferring to a human |
| Satisfaction | Aggregate score based on call outcomes and sentiment |
Use the date picker in the top-right corner to change the reporting period. Compare week-over-week to spot trends.
Call Status Meanings
Every call in your log has one of these statuses:
- Completed - The AI handled the call from start to finish. The caller got the information they needed or booked an appointment.
- Transferred - The AI could not fully help the caller and transferred them to you or your team. Review these transcripts to find knowledge base gaps.
- Missed - The call rang but was not answered. This can happen if the system is temporarily at capacity or if the caller hangs up before the AI connects.
- Failed - A technical issue prevented the call from connecting. If you see multiple failed calls, contact support.
- Voicemail - The caller chose to leave a message instead of speaking with the AI.
AI Confidence Scores Explained
Each call transcript includes a confidence score from 0 to 100:
- 90 - 100: The AI was highly confident in its answers. No action needed.
- 70 - 89: The AI answered correctly but had to work harder to find the right information. Consider adding more detail to related knowledge base articles.
- 50 - 69: The AI was uncertain about some answers. Review the transcript and fill in knowledge gaps.
- Below 50: The AI struggled significantly. The caller likely did not get useful answers. Prioritize updating your knowledge base for this topic.
Tip: Filter your call log by confidence score to quickly find calls that need attention.
Using Sentiment Analysis
AIVO analyzes the tone and language of each call to assign a sentiment:
- Positive - Caller seemed satisfied, friendly, or grateful.
- Neutral - Standard informational call with no strong emotion.
- Negative - Caller seemed frustrated, confused, or unhappy.
Use sentiment data to:
- Identify topics that consistently frustrate callers.
- Spot training gaps (negative sentiment + low confidence = missing knowledge).
- Celebrate wins (high positive sentiment means your AI is doing a great job).
Exporting Reports
- Navigate to Calls or Analytics in your dashboard.
- Set your desired date range and filters.
- Click the Export button in the top-right corner.
- Choose your format: CSV for spreadsheets or PDF for reports.
- The file downloads immediately.
Exported CSV files include: date, time, caller number, duration, status, confidence score, sentiment, summary, and topics discussed.
Was this article helpful?