Campaign live monitoring (scheduled / running / paused)

Live monitoring is the operator view. It is where you detect problems early, understand blockers, and decide whether to adjust, pause, or let it run.

At a glance

• KPI strip shows progress, task states, answer rate, duration, cost, and goal/budget signals.
• Dispatch and queue health explain why tasks are pending.
• Tasks table is where you see blocker codes, retries, and debug entry points.

KPI strip (what it means)

Expect to see:

• progress
• in progress / completed / failed / pending task counts
• answer rate
• average duration
• total cost
• optional budget progress
• optional goal progress

How to interpret KPI shifts

• If pending grows while in progress stays flat: you’re usually capacity blocked (numbers, windows, queue health).
• If failed grows quickly: inspect blocker codes and provider errors before you scale.
• If answer rate drops: check number pool health and country match assumptions.

Paused state

Paused campaigns display a banner with pause time and pause reason.

Monitoring sections (dashlets you should read)

Campaign timeline

Time-series view of campaign activity.

Dispatch & queue health

The “why is it not dialing?” panel. Use this before changing random settings.

Typical operator checks:

• Are we outside call windows?
• Is concurrency effectively zero (tenant cap or pool capacity exhausted)?
• Are tasks blocked by retry timing (next_eligible_at)?
• Are there provider-level failures or throttling?

Disposition breakdown

Distribution of outcomes. Use it to detect script mismatch or audience mismatch.

Sentiment

High-level quality indicator. Don’t over-trust it — use it as a pointer to investigate.

Retry effectiveness

Shows whether retries are helping or just increasing spam risk.

A/B test results

Variant performance signals when A/B is enabled.

Cost summary

Cost signals for governance and scale decisions.

Tasks table (where debugging happens)

The tasks table shows per-contact task state and blockers.

Operators should understand:

• task status
• blocker codes
• attempts vs max attempts
• system retries
• duration
• score
• last attempted

Task table columns (what each means)

• Contact: who the task is for (link back to contact context).
• Phone: destination number.
• From: from-number used (or sticky number).
• Status: current task state.
• Blocker: why it can’t progress right now (if blocked).
• Disposition: outcome label (when completed).
• Attempts: attempts already used.
• System retries: system-level retry scheduling count.
• Duration: talk time (when applicable).
• Score: call quality/score (when applicable).
• Last attempted: last attempt timestamp.
• Debug: deep inspection entry point.

Debug entry point

Use the debug view when:

• a task is stuck
• a blocker is unclear
• provider errors occur

What to capture before escalating:

• a screenshot of the tasks table row (status + blocker)
• the debug view evidence (events/snapshots)
• the campaign’s dispatch health card

Health sections

Number pool health

Shows answer rate and weight by number. Use it to identify bad numbers early.

Operator moves when answer rate craters:

1. Remove the worst-performing number from the pool.
2. Reduce concurrency for containment.
3. Narrow windows to supervised times.
4. Re-check country match logic (if enabled).

Staged CSV audience upload

When a campaign has a staged audience file, a CSV metadata card may appear.