Outbound Metrics

Metric Name

The recognized name, e.g., First Response Time (FRT)

Metric Purpose

What aspect of performance does it measure

Measurement Unit

Time (seconds/minutes), count, percentage, score, etc.

Calculation Method

Formula or approach (e.g., FRT = time from customer contact to first agent response)

Context/Scope

Which channels, agent groups, or periods does the metric apply to

Benchmark/Standard

Industry typical target or range values

Counter

A cumulative metric that represents a single monotonically increasing counter whose value can only be incremented or reset to zero on restart. Use for counts of events (e.g., successful dials, total calls).

Gauge

A metric that represents a single numerical value that can arbitrarily go up and down. Use for current states (e.g., contacts remaining, number of active agents).

Histogram

Samples observations (usually durations or sizes) and counts them in configurable buckets. It also provides a sum of all observed values. Use for request latencies or durations (e.g., wait time, wrap time).

Summary

Similar to a Histogram, but also calculates configurable client-side quantiles over a sliding time window. Use for service-level objective (SLO) monitoring of durations.

Recently

between the last 5 minutes and the last 30 days

Interval

The sampling interval can be any time duration between 1s and 15 m.

Entity (Object State & Flow) Metrics

Monitors the aggregate state and flow of core business entities (e.g., Contacts, Campaigns, Agents). Metrics track the population count residing in specific states (Inventory/Backlogs via Gauges) and the cumulative volume that has completed transitions (Throughput/Outcomes via Counters).

• e.g., ob_contacts_to_load (Gauge): The current count of contacts awaiting processing (Inventory/Backlog).

• e.g., ob_contacts_loaded_total (Counter): The cumulative number of contacts that have completed the loading step (Throughput).

Flow Metrics

Monitors the real-time operational values and dynamic performance ratios generated at key decision and control points within the Outbound Dialing Flow. These metrics are crucial for controlling the system's behavior (e.g., dialing mode/efficiency, agent reservation, cpa logic, error rate, etc. ) and optimizing the flow's moment-to-moment efficiency. They represent instantaneous calculated states or ratios, and do not track the aggregate status or cumulative outcome of contact entities.

• e.g., ob_cpa_machine_ratio (Gauge): The instantaneous ratio of calls currently classified as Machine vs. Human by the Call Progress Analysis (CPA) node. (A real-time performance indicator of the flow's effectiveness).

• e.g., ob_init_calls_per_agent (Gauge): The calculated value used by the INIT Node that determines the rate of call initiation per every reserved agent, which dictates the dynamic dialing mode (predictive vs. power). (A real-time flow control parameter).

Performance Metrics

Measures the time taken to complete critical system actions and state transitions (Service Level Indicators).

• e.g., api_fetch_duration_seconds_p99 (Histogram/Summary): The 99th percentile of time taken to fetch a contact record from the database (System Responsiveness).

• e.g., agent_handle_time_seconds_avg (Gauge): The average time an agent spends on a single connected call (Process Speed).

ob_contacts_to_load

PLANNED

Gauge

The number of contacts pending (from CX Filters) to be ingested into this campaign.

Allows you to monitor:

• The backlog and speed of the contact ingestion process.

• campaignName

• High (Fine): 1 Minute

Campaign Init

ob_contacts_loaded_total

AVAILABLE

Counter

The cumulative number of unique (unless a restart occurs) contacts successfully ingested and made available for dialing/contacting in this campaign flow.

Allows you to monitor:

• The total size of the contact list and verify ingestion completion.

• campaignName

• order (FIFO/LIFO)

• High (Fine): 1 Minute

Campaign Init

ob_contacts_unqualified_total

PLANNED

Counter

The cumulative count of contacts scrubbed/removed from the list before any dialing attempt.

Allows you to monitor:

• List data quality and loss volume.

• campaignName

• reason

  • (e.g., format_issue, duplicate)

• Intermediate (5-10 Minutes)

Dialer Nodes

• CCX Dialer

• CCE Dialer

• CX Dialer

ob_contact_condition_evaluations_total

PLANNED

Counter

The cumulative count of contacts that have been evaluated against a specific condition set up within the campaign flow.

Allows you to monitor:

• The throughput and success rate of a specific business rule or segment filter within the execution flow.

• campaignName

• condition_name

  • The name of the specific condition node, e.g., "CountryCheck"

• status

  • passed or failed

• Intermediate (5 Minutes)

Condition

ob_contact_decision_evaluations_total

PLANNED

Counter

This metric counts the number of contacts that have passed a specific condition within a campaign's decision node.

Allows you to monitor:

• The volume and effectiveness of contact recent filtering to ensure only eligible contacts continue in the campaign flow.

• campaignName

• decision_name

  • The name of the specific decision node, e.g., "RetryOnBusy"

• status

  • passed or failed

• reasonCode

  • e.g., busy, read

• Intermediate (5 Minutes)

Decision

Dialer Metrics

ob_contacts_sent_for_dialing

AVAILABLE

Counter

This metric counts the total number of contacts pushed to the Dialer Nodes for execution. It can be broken down for more granular insights:

I would like to monitor:

• number of ob_contacts sent for dialing every execution cycle

• total ob_contacts sent for dialing so far (per channel)

• dialing volume by hour, i.e. number of outbound contacts sent for dialing in each hour of the day to compare campaign dialing volume trends by time

• campaignName

• channelType

5 min

UCCE/UCCX Dialer, EFCX Dialer

ob_call_results

AVAILABLE

Counter

This metric counts the total number of recent call results generated by the dialer. It can be broken down for more granular insights:

I would like to monitor:

• the total number of recent call results, segregated by outcome, so that I can analyse performance effectively.

• call results segmented by status and reasonCode, so that I can perform a more detailed analysis.

• campaignName

• channelType

• status (high-level result: answered, busy, etc)

• reasonCode (user_declined

5 min

UCCE/UCCX Dialer, EFCX Dialer

ob_dialing_in_progress

AVAILABLE

Gauge

This metric offers a live snapshot of your dialer's workload by counting the number of in-progress contacts currently being dialed.

I would like to monitor:

• I would like to monitor dialer load in real time.

• I would like to track dialing volume versus available agents so that I can help avoid overload.

• campaignName

• channelType

5 min

UCCE/UCCX Dialer, EFCX Dialer

Digital Messaging Metrics

ob_messages_sent_total

AVAILABLE

Counter

This metric counts the number of outbound messages sent per campaign and per channel type, allowing you to track the total message volume across all digital channels, like SMS and email. Key uses include:

• Channel Load Monitoring: Identify if certain channels are handling disproportionately high volumes, which may indicate overuse

• Campaign Activity Tracking: Confirm that campaigns are actively sending messages as expected

• Historical Trend Analysis: Track message volume trends over time to evaluate seasonality, growth, or decline in outreach

• Quota and Cost Monitoring: Estimate costs or monitor message limits, especially in environments with channel-based pricing (e.g., SMS)

• campaignName

• channelType

5 min

Digital Channel Message

ob_message_results_total

AVAILABLE

Counter

This metric counts the number of outbound messages that came back with a message result per campaign and channel type, allowing you to track the delivery performance of digital messaging campaigns, investigate failures by channel or campaign, and refine strategy by identifying high-performing channels.

Key uses include:

• Measure the effectiveness and delivery success rates of your digital messaging campaigns.

• Diagnose delivery problems by filtering for failed results on specific channels or campaigns.

• Optimize your channel strategy by identifying which channels (channelType) provide the highest delivery success rates.

• campaignName

• channelType

• result

  • e.g., Failed/ Sent

5 min

Digital Channel Message

ob_message_delivered_total

ROADMAP

Counter

Counts the total number of digital outbound messages (SMS, WhatsApp, Email, etc.) that were successfully delivered to recipients, as confirmed by the provider.

1. Delivery Assurance: Verifies how many outbound messages reached the intended recipient.

2. Channel Performance Tracking: Helps compare delivery success across channels.

3. Failure Diagnosis Support: When paired with failure metrics, helps identify drop rates, blocks, or provider issues.

I would like to monitor:

• Number of messages successfully delivered per campaign and channel

• Delivery rate per channel: (message_delivered_total ÷ message_sent_total) × 100

• campaignName

• channelType

5 min

Digital Channel Message (via provider delivery APIs)

ob_message_failed_total

ROADMAP 2026

Counter

Counts the number of outbound digital messages (SMS, WhatsApp, Email, etc.) that failed to deliver due to issues like invalid numbers, content rejections, or provider/network errors.

• Failure Tracking: Shows how many messages did not reach the end user.

• Channel Quality Control: Helps detect which channels have higher error rates.

• Diagnostic Signal: Indicates list quality, opt-out enforcement, or content/template issues.

I would like to monitor:

• Total number of failed messages per campaign and channel

• Common failure types: invalid number, blocked content, opt-out violation

• Channels with high failure rates to improve targeting or compliance

• campaignName

• channelType

• failureReason (e.g. WhatsApp: template_rejected, unregistered_number, rate_limited, session_expired, provider_error, SMS: carrier_blocked, unsupported_country, Email: spam_blocked, hard_bounce, soft_bounce…)

5 min

Digital Channel Message (via provider delivery APIs)

ob_message_read_total

ROADMAP

May be available for limited channels.

Counter

Counts the number of outbound digital messages (e.g., WhatsApp, Messenger) that were successfully read or seen by the recipient, based on read receipts supported by the messaging platform.

• Engagement Tracking: Measures the level of actual message interaction, beyond just delivery.

• Channel Performance Indicator: Helps compare user engagement across different digital channels.

• Supports Optimization: Enables refining retry and fallback logic based on whether users are actually reading messages.

I would like to monitor:

• How many messages per campaign were read after delivery

• Compare engagement across WhatsApp, Messenger, etc.

• Identify low-engagement channels or message formats

• campaignName

• channelType

5 min

Digital Channel Message (via provider delivery APIs)

ob_message_read_rate

ROADMAP

May be available for limited channels.

Gauge (Calculated)

Measures the percentage of delivered messages that were actually read/opened (where supported by the channel).

1. Engagement Indicator: Reading behavior signals contact interest, not just delivery.

2. Performance Benchmarking: Helps determine which channel generates more engagement

I would like to monitor:

• percentage of messages read per channel or campaign

• Compare engagement between WhatsApp, SMS, email, etc.

• Improve fallback logic based on read behaviour

Formula

• (ob_message_read_total ÷ ob_message_delivered_total) × 100

• campaignName

• channelType

5 min

Digital Channel Message (via provider delivery APIs)

ob_agents_seized_gauge

AVAILABLE

DEV-PLANNED for the addition of the queueName label.

Gauge

The current number of agents actively reserved ("seized") by the system for the campaign's current dialing requirements. This is a real-time count.

This metric directly reflects the success of the pacing algorithm in reserving the necessary human resources calculated by the Seize Node.

Allows you to monitor:

• Real-time agent reservation to ensure the dialer has the required resources to match the dynamic dialing ratio.

• The success of the pacing algorithm in reserving resources. Observing this value against ob_agents_needed_gauge (the calculated requirement) shows if the Seize Node is actively meeting the need or is over-seizing/under-seizing.

• campaignName

• queueName

10 Seconds (or less)

SIEZE

ob_agents_available_gauge

ROADMAP 2026

Gauge

The total number of agents currently logged in, ready, and potentially available for seizing by any outbound campaign in the Contact Center (CC) that utilizes the specified Queue. This represents the total capacity pool.

Allows you to monitor:

• The overall available capacity of the Contact Center for outbound work, segmented by skill or queue.

• Capacity shortfalls: Comparing this metric to the outbound_agents_seized_gauge metric identifies if the outbound demand is exhausting the total agent pool, triggering a staffing alert if the remaining capacity is too low.

• queueName

1 min

SIEZE

ob_agents_idle_gauge

ROADMAP 2026

Gauge

The current number of agents who are logged in, available, and are currently in an idle/waiting state (i.e., not seized and not on a call) and available to be seized for outbound work.

Allows you to monitor:

• The immediate surplus capacity that is not yet utilized by the dialer.

• Agent Utilization: A persistently high idle count indicates the dialer is not aggressive enough or the list is exhausted, leading to low utilization.

• queueName

1 min

SIEZE

ob_agents_acw_gauge

ROADMAP 2026

Gauge

The current number of agents actively performing After Contact Work (ACW), which is post-call wrap-up, data entry, etc. These agents are temporarily unavailable for new calls.

Allows you to monitor:

• Resource recovery timing. These agents represent the next wave of available resources.

• ACW adherence: A high or fluctuating count may signal that agents are not using ACW time efficiently, impacting overall productivity.

• queueName

1 min

SIEZE

ob_agent_talk_time_seconds_sum

ROADMAP 2026

Counter

The cumulative sum of time (in seconds) spent by agents talking to connected contacts within the campaign flow. Used with the ob_connected_calls_total to calculate the Average Talk Time (ATT).

Allows you to monitor:

• Agent Productivity: Track agent-level talk time volume (quantity).

• Pacing Algorithm Input: The historical average talk time is a critical component for predicting when an agent will become free.

• campaignName

• agentId

5 min

Dialer Nodes

ob_dial_ratio_setting

DEV-PLANNED

Gauge

The current contact ratio (contact per agent siezed) is currently configured or dynamically calculated for this campaign.

Allows you to monitor:

• Track the dialing mode and confirm the dynamic calculation logic is functioning correctly.

• campaignName

5 Seconds (or less)

INIT

ob_cpa_result_count

DEV-PLANNED

Counter

• campaignName

• result

  • human

  • answering machine

  • failed

Dailer Node

ob_cpa_machine_ratio

DEV-PLANNED

Gauge (Calculated)

The instantaneous ratio of calls currently classified as Machine vs. Human by the Call Progress Analysis (CPA) node within the outbound flow. A real-time performance indicator of the flow's effectiveness. The value ranges from 0.0 to 1.0.

Allows you to monitor:

• Dynamic Pacing Adjustment: Provides real-time input for the predictive dialing algorithm. Change Indication: A rising ratio signals the pace must slow down; a falling ratio signals the pace can safely increase.

• Instant List Quality Check: Sustained high ratio -> Current list segment has poor answer quality (high voicemail rate).

• SLA and Efficiency: A sharp spike triggers an alert, indicating an immediate need for operational review to prevent an agent idle-time event.

Formula

• ob_cpa_result_count{human/ansering_machine}

• campaignName

10 Seconds (or less)

CPA

ob_pending_sieze_requests_total

ROADMAP

Gauge

• campaignName

• queueName

ob_agents_needed_gauge

ROADMAP

Gauge

The real-time number of agents that the Pacing/INIT Node has determined are immediately required to prevent connected contacts from being abandoned. This value represents the current human resource demand of the flow.

Allows you to monitor:

• Demand Signal Health: Tracks the instantaneous need created by the current rate of answered calls that require an agent connection.

• Resource Gap Analysis: Used in conjunction with ob_agents_seized_gauge to measure the gap (shortage or surplus) between the calculated need and the agents actually reserved by the flow.

• Abandoned Call Risk

• campaignName

• queueName

5 secs

INIT

ob_abandoned_calls_total

DEV-PLANNED

Counter

Counts outbound calls that were abandoned before reaching an agent or IVR — whether the contact hung up during ringing or the dialer failed to connect an agent on time.

I would like to monitor:

• total outbound abandoned calls per campaign

• breakdown by True vs Compliance abandon types

• hourly abandonment rate to detect agent capacity constraints or pacing issues

• campaignName

UCCE/UCCX Dialer or EF Voice Dialer,

derived from ob_call_results

ob_connected_calls_total

ROADMAP 2026

Counter

The cumulative number of successful outbound calls that were answered by a human and connected to an agent in the campaign flow.

Allows you to monitor:

• Agent Throughput (Quantity): The total volume of productive interactions handled by an agent or campaign.

• Average Talk Time (ATT) Denominator: This metric serves as the count for calculating ATT (when divided into ob_agent_talk_time_seconds_sum).

• Hit Rate/Conversion Rate: Forms the base for calculating the percentage of calls that lead to a business outcome or sale.

• campaignName

• agentId

5 min

Dialer/CPA Node

ob_dial_to_connect_duration_seconds

ROADMAP 2026

Histogram / Summary

The duration (latency) from the moment the dialer initiates the call to the moment a live person answers (i.e., CPA is confirmed human).

Allows you to monitor:

• Network Latency and List Responsiveness: Measures how quickly calls are answered, which is key for accurate predictive timing.

• Service Objective: A primary SLI for the quality of the dialing infrastructure. The target P95 should be set (e.g., < 4.0 seconds).

• campaignName

• result (e.g., success, timeout)

Per Call Event

Dialer/CPA Node

ob_hold_time_seconds_avg

ROADMAP 2026

Gauge / Summary

The average time a successfully answered contact spends on the line waiting for a seized agent to become available. (Also known as "Agent Wait Time" for the customer).

This is the most critical compliance and customer experience SLI for predictive dialing. It is regulated by law (ACR, which is directly impacted by high hold time).

Allows you to monitor:

• Customer Experience SLI: A measure of the dialing system's over-aggressiveness. High latency here directly leads to high abandoned calls.

• Pacing Health Check: A high average indicates the dialer ratio (ob_dial_ratio_setting) is set too high for the current agent availability.

• campaignName

• queueName

10 secs

SIEZE/ INIT

ob_amd_detection_errors_total

ROADMAP 2026

Counter

The cumulative number of calls where the Call Progress Analysis (CPA/AMD) node incorrectly classified the answer (e.g., mistook a human for a machine, or vice versa).

Allows you to monitor:

• CPA Reliability SLI: An indicator of the technical quality of the answering machine detection (AMD) algorithm.

• Compliance Risk: A high error rate (especially failing to detect a machine) impacts agent efficiency and list utilization.

• campaignName

• detectionType (e.g., Human-to-Machine Error)

5 mins

CPA Node

ob_contact_rate_ratio

ROADMAP 2026

Core Campaign Effectiveness (Highest Importance): This is the ultimate metric for list quality and dialer performance combined. A low rate means the dialer is wasting resources (time, lines) on non-productive dials (busy, no answer, disconnected). It must be constantly monitored to ensure the list is viable and the campaign is economically efficient.

Gauge (Calculated)

The instantaneous ratio of calls resulting in a live human answer (Right Party Contact or any live person) compared to the total number of dials made.

Allows you to monitor:

• Real-Time List Quality: A low contact rate signals poor list data (e.g., disconnected numbers, high volume of non-human answers).

• Overall Dialer Effectiveness: Confirms that the dialer is successfully reaching humans

• Economic Efficiency: Helps determine if the cost of placing dials is justified by the return in live connections.

Formula

• ob_connected_calls_total / ob_dials_total

• campaignName

10 secs

Calculated from Dialer/ CPA Nodes

ob_dropped_call_rate_ratio

ROADMAP 2026

This high frequency is mandatory for maintaining regulatory compliance, as the abandoned call rate must be continuously tracked and controlled to stay below a legally mandated threshold (typically 3% or 5%).

Gauge (Calculated)

The instantaneous ratio of calls that were answered by a contact but dropped (abandoned) before being connected to an available human agent or in IVR.

Allows you to monitor:

• Regulatory Compliance SLI: Tracks adherence to laws governing predictive dialing aggressiveness (e.g., Federal Trade Commission rules).

• Customer Experience Impact: A high rate signals over-aggressiveness, leading to poor customer experience and potential complaints.

• Pacing Algorithm Health: Directly shows when the dialer's predictions about agent availability are incorrect.

Formula

• ob_abandoned_calls_total / ob_answered_calls_total

• campaignName

• abandonType (true / compliance)

• AbandonReasonCode (Dialer Abandon, Abandon to IVR, Customer Abandon)

10 secs

Calculated (from Dialer/INIT)

ob_list_penetration_ratio
ROADMAP 2026

This metric tracks the completeness of the dialing effort. It helps managers understand how close the campaign is to being finished and whether the dialer is stuck, cycling on a small subset of records, or ignoring parts of the list. It is crucial for deciding when to stop the campaign or transition to a different dialing mode (e.g., from predictive to power).

Gauge (Calculated)

The ratio of the number of unique contacts that have been dialed (attempted) at least once, compared to the total number of contacts loaded into the campaign.

Allows you to monitor:

• Campaign Completion Progress: Tracks the percentage of the contact list that has been engaged by the dialer.

• List Health and Stuck Records: A low penetration rate over an extended period indicates the dialer is cycling repeatedly on a small subset of records or failing to access the rest of the list.

• Management Decision Point: Informs managers when to terminate the campaign, change the list, or switch to a less aggressive dialing mode for the remaining records.

Formula:

• ob_unique_contacts_dialed / ob_contacts_loaded_total

• campaignName

1 hour

This metric is best aggregated over a campaign batch or hour, as it represents progress toward list exhaustion rather than real-time flow control.

INIT