Understanding Results

Calls Analysis Results

Calls Analysis results provide a qualitative evaluation of each call, including:

Topics — Main subjects discussed during the call
Agent Mistakes — Errors or protocol violations identified
Missing Prompts — Required protocol steps the agent skipped
Goal Reached — Whether the call objective was achieved, with explanation

Results are displayed in a structured view with categories and supporting quotes from the transcript.

COPC Evaluation Results

COPC Evaluation results are displayed in two tabs: Dashboard and Table.

Dashboard Tab

The Dashboard provides a high-level overview of evaluation results across all scored calls.

KPI Cards at the top show:

Metric	Description
Total Calls	Number of calls evaluated
Average Score	Mean score percentage across all calls
Critical Errors	Count and percentage of calls with critical errors
Passing Rate	Percentage of calls scoring 80% or above without critical errors

Score by Criterion — A bar chart showing the average score for each criterion across all calls. Each bar displays the average points and percentage achieved.

Score Range — Shows the minimum, average, and maximum scores across all evaluated calls.

Table Tab

The Table view lists every evaluated call with columns for:

File — Name of the call recording
Score — Overall percentage score (color-coded)
Critical — Whether a critical error was detected
Topic — Main topic of the call
Sentiment — Overall call sentiment (positive, neutral, or negative)

Columns can be sorted by clicking their header.

If the project has call metadata, the table also shows Agent, Call date, Direction, Caller number and Campaign, and gains filters for agent, direction and a call-date range. These columns and filters appear only when the underlying data exists, so a project without a manifest looks exactly as before.

Expanding a row reveals the detailed criterion-by-criterion breakdown:

Each criterion shows score achieved vs. maximum points
A progress bar visualizes the score
Justification — AI-generated explanation of why the score was assigned
Quote — Direct excerpt from the transcript supporting the evaluation. Click the quote’s transcript icon to open the transcript at that moment and check the evaluation against what was actually said
Critical error criteria are flagged with a warning icon
Criteria that did not apply to this call carry an N/A badge, with their points in grey; a line above the table reports how many were skipped

Merged Results

With automatic analysis enabled, calls are analyzed in small batches as they arrive. Rather than listing every batch, the Results tab shows one merged result per analysis type, marked all analyzed calls, that covers every call analyzed so far. It opens and exports like any other result.

Free-text output (topics, mistakes, missing prompts, custom-question answers) is grouped into categories. New calls are classified against the categories already in place; the Refresh categories button on a merged result re-groups everything from scratch in the background, and a daily sweep does the same automatically.

Per-Call vs Cross-Call

A result document answers questions about the calls in it. For the whole operation — trends over time, contact reason mix, day versus night, agent league tables, and a searchable list of every analyzed call — use the Call Center Report.

Score Color Coding

Scores are color-coded throughout the results:

Score Range	Color	Meaning
80% – 100%	Green	Meets expectations
60% – 79%	Yellow	Needs improvement
Below 60%	Red	Below standards
Critical error	Red	Automatic 0% — call fails regardless of other scores

Critical Error Behavior

When a call triggers a critical error:

The criterion marked as critical received a score of 0
The entire call score is set to 0%
The call is flagged with a warning indicator in the results table
The expanded detail shows which criterion triggered the failure
The call counts toward the critical error statistics in the Dashboard

This ensures that non-negotiable quality requirements (e.g. identity verification, compliance disclosures) are strictly enforced.

Criteria That Did Not Apply

Not every criterion applies to every call. When the situation a criterion describes never arose, or the call ended before the agent reached that step, the criterion is skipped and awarded its full points.

A skipped criterion carries an N/A badge, and its points are shown in grey rather than green — they still count toward the call total, so the arithmetic adds up either way. A line above the expanded table reports how many of the call’s criteria were skipped, and each justification says which condition was absent or where the call ended.

Two things follow, and both catch people out:

A very short call can score 100%. If the customer hung up during the greeting, almost nothing the agent owed had come due yet. The score is correct.
A conditional criterion reads high in the averages, because the calls where it did not apply are counted at full marks. The Score by Criterion chart and the Call Center Report scorecard both report how often the criterion actually applied, so you can see when that is what is happening.

See How Scoring Works for what the evaluator skips, what it never skips, and how to read a criterion average that includes skipped calls.

Exporting Results

Select one or more completed analysis tasks and click Export:

Word (.docx) — Formatted report with all scores and justifications
Excel (.xlsx) — Structured data for further analysis and filtering
PowerPoint (.pptx) — Presentation-ready slides with charts

For a client-ready QA document with an executive summary, findings and recommendations across all calls in a period, use Export PDF in the Call Center Report.