Call Center Analytics

Overview

Call Center Analytics gives you a focused view of your human queue's performance — queue wait times, agent handle times, abandon behavior, and bot effectiveness. It is designed specifically for voice and chat contact center operations, where queue dynamics and handoff quality directly affect customer satisfaction.

To access it, go to Reports → Call Center Analytics in the left sidebar. The report defaults to the last 7 days and can be filtered by date range, team, and channel.

Call Center Analytics tracks the full journey of a conversation: how long a bot handled it before handing off, how long the customer waited in the human queue, how long the agent worked on it, and whether the customer stayed or abandoned. These dimensions are measured separately so you can pinpoint exactly where time is lost.

---

Key Metrics Explained

Avg Handle Time (AHT)

Time from the agent's first reply to conversation close. This measures pure agent work time — it excludes the time the customer spent waiting in the queue and any time the bot handled the conversation before the handoff. AHT reflects how efficiently agents resolve issues once they engage.

Avg Queue Wait

Time from when the conversation entered the human queue to when the agent first replied. Bot handling time before the handoff is excluded. This is the metric most directly linked to customer frustration — long queue waits drive abandonment.

Avg Service Time

Queue wait + handle time combined. This represents the total agent-touched time in a conversation — from the moment the conversation became a human concern to the moment it was closed.

Avg Resolution Time

The full end-to-end duration: from conversation open (including any bot preamble) to close. This is the number customers experience, not the number agents control. Bot preamble is included here, which is why resolution time is typically longer than service time.

Bot Preamble Time

Time the bot handled the conversation before it entered the human queue. Comparing preamble time to handle time tells you how much work bots are absorbing upstream of your agents.

Bot Containment Rate

Percentage of bot-started conversations fully resolved by the bot — meaning the conversation never entered the human queue and was closed cleanly. A bot that times out, gets abandoned, or transfers to a human does not count as contained. This is a strict definition by design: inflating containment with partial resolutions obscures gaps in bot effectiveness.

Bot-to-Human Transfer Rate

Percentage of bot-touched conversations that transferred to a human agent. The complement of containment rate for conversations that did not self-resolve. Use this alongside containment rate to understand bot routing behavior.

Abandon Rate

Percentage of queued conversations where the customer left voluntarily before an agent responded. This is a customer-side metric — the customer chose not to wait. Elevated abandon rates typically signal queue wait times that exceed customer tolerance.

Missed Rate

Percentage of queued conversations the system failed to answer — rejected or timed out on the agent side. This is an agent-side metric and is distinct from Abandon Rate. A high missed rate suggests staffing gaps or routing misconfiguration, not customer impatience.

Critical Abandon Rate

Percentage of queued conversations abandoned after waiting more than 2 minutes. This subset of Abandon Rate isolates genuine customer frustration — customers who waited well past typical patience thresholds. Critical Abandon Rate is a leading indicator of satisfaction problems before they surface in CSAT scores.

---

Abandon Depth

The Abandon Depth chart shows at what wait time customers are abandoning, broken into five buckets:

Bucket	Range
Immediate	Under 30 seconds
Short wait	30–60 seconds
Medium wait	1–2 minutes
Long wait	2–5 minutes
Extended wait	5+ minutes

How to use this chart: If most abandons happen in the 2–5 minute bucket, your SLA target should be under 2 minutes — customers are willing to wait that long but lose patience beyond it. If you see a spike in the "under 30 seconds" bucket, it may indicate customers who hit the queue, saw no progress indicator, and left immediately — a UX or queue messaging issue rather than a staffing issue.

Abandon Depth is most useful when evaluated alongside Critical Abandon Rate and Avg Queue Wait to set realistic, data-backed SLA targets for your teams.

---

Bot Analytics Tab

The Bot Analytics tab provides a dedicated view of bot performance within the call center context:

• Conversations handled by bot — total volume the bot engaged with
• Bot Containment Rate — percentage resolved by the bot without human handoff (strict definition: never entered the human queue, closed cleanly)
• Bot-to-Human Transfer Rate — percentage that transferred to an agent
• Avg Bot Preamble Time — average time the bot spent before handing off

Why the strict containment definition matters: A bot that runs out of retries, gets abandoned, or times out while technically "in control" is not a successful containment — the customer's issue was not resolved. Counting those as containment inflates the metric and gives a misleading picture of bot effectiveness. Velaro uses the strict definition so the number reflects genuine self-service success.

Use the Bot Analytics tab to evaluate whether your bot is reducing queue load or adding friction. If containment is low and preamble time is high, the bot may be delaying customers without resolving them — a common sign that bot knowledge needs expansion.

---

Advanced Reports: Teams and Channels

The Team Breakdown and Channel Breakdown tabs show how key metrics vary across your teams and conversation channels (web, SMS, WhatsApp, Facebook, etc.).

What's included by default vs. what requires an add-on:

Capability	Availability
Overview metrics (volume, speed, abandon, bot)	Included in all plans
Threshold alerts	Included in all plans
Team breakdown	Advanced Analytics add-on
Channel breakdown	Advanced Analytics add-on
Agent breakdown	Advanced Analytics add-on
Email digest reports	Advanced Analytics add-on

Team and channel analytics, agent-level reporting, and email digest capabilities may require an add-on subscription. Contact your account manager or our sales team for details on what's included in your current plan. Moshky can confirm availability for your specific account.

---

Threshold Alerts

The Alerts & Thresholds tab lets you set warning and critical thresholds for key metrics. When a metric breaches a threshold, the alert is surfaced in the Velaro UI so supervisors can act before the situation worsens.

Default thresholds:

Metric	Warning	Critical
Avg Queue Wait	90 seconds	3 minutes
Abandon Rate	10%	25%
Critical Abandon Rate	5%	15%
Missed Rate	5%	15%

Thresholds are configurable. The defaults are calibrated to typical contact center benchmarks — adjust them to match your team's SLA commitments and customer expectations.

Alerts are not yet delivered by email or SMS — they appear in the Reports dashboard. If you need proactive alerting, consider pairing threshold monitoring with a scheduled Moshky check.

---

Email Digest Reports

Email digest reports deliver a scheduled summary of your call center KPIs directly to your inbox. Each digest includes:

• Key metric summary for the period (volume, AHT, queue wait, abandon rate, bot containment)
• Trend comparison vs. the prior equivalent period
• Any threshold breaches that occurred during the period

Digests can be scheduled weekly, bi-weekly, or monthly. They are sent to the email addresses you configure in the digest settings tab.

Team and channel analytics, agent-level reporting, and email digest capabilities may require an add-on subscription. Contact your account manager or our sales team for details on what's included in your current plan. Moshky can confirm availability for your specific account.

---

How These Metrics Differ from Historical Reports

Call Center Analytics and Velaro's Historical Reports both measure conversation performance, but they answer different questions — and the numbers will differ.

Dimension	Call Center Analytics	Historical Reports
Scope	Human-queue conversations only	All conversations, all channels
Baseline	Queue entry time	Conversation StartTimestamp
Async channels (Email, Ticket)	Excluded	Included
Bot preamble	Measured separately	Included in resolution time

The key difference: Call Center Analytics uses the moment a conversation enters the human queue as its baseline. Historical Reports use the conversation start timestamp, which includes any bot handling that occurred before a human was involved.

This means the same conversation will show a shorter "queue wait" in Call Center Analytics than it shows as "first response time" in Historical Reports — both numbers are correct, they measure different intervals. Call Center Analytics is optimized for evaluating queue and agent performance; Historical Reports are optimized for overall volume trends and cross-channel comparison.

Neither report is more accurate — they answer different operational questions. Use Call Center Analytics when you are managing staffing, SLAs, and queue health. Use Historical Reports when you are analyzing total volume, channel mix, or long-term trends.

---

FAQ

Q: Why does my Bot Containment Rate look lower than expected?

Velaro uses a strict containment definition: the conversation must never have entered the human queue, and it must have been closed cleanly. Conversations where the bot timed out, the customer abandoned, or the bot transferred to a human are not counted as contained. This is intentional — it keeps the metric honest. If containment is lower than expected, check bot preamble time and transfer rate to identify where conversations are falling through.

Q: What is the difference between Abandon Rate and Critical Abandon Rate?

Abandon Rate counts every conversation where a customer left the queue before an agent responded — including customers who waited only a few seconds. Critical Abandon Rate counts only those who waited more than 2 minutes before leaving. Critical Abandon Rate is a more reliable signal of genuine frustration; it filters out customers who queued and immediately changed their minds.

Q: Why does Avg Resolution Time include bot preamble, but Avg Service Time does not?

Resolution Time measures the full customer experience — from the moment they first interacted to the moment the conversation closed. Bot preamble is part of that experience. Service Time measures only the agent-side work (queue wait + handle time). Separating them lets you see exactly how much time bots contribute upstream of your agents.

Q: My Call Center Analytics numbers are different from my Historical report. Which one should I trust?

Both are correct — they measure different things. Call Center Analytics uses queue entry time as its baseline and excludes async channels (Email, Ticket). Historical Reports use conversation start time and include all channels. The same conversation will produce different numbers in each report. Use Call Center Analytics for queue management and SLA decisions; use Historical Reports for volume trends and cross-channel analysis.