Number Health

This guide is for anyone who owns outbound voice performance at scale — voice AI operators deploying AI agents or blended AI-plus-human call flows, outbound sales and SDR/BDR managers running large calling programs, contact center and CX operations managers in collections, retention, or service environments, and telephony and RevOps practitioners who configure dialers and manage number pools.

If you manage phone numbers (DIDs) as part of a voice operation — with human agents, AI agents, or both — this guide covers how to measure number health, act on it faster, and build systems that surface problems before they surface in your answer rate.

Experienced outbound teams rotate numbers, warm up fresh DIDs, segment pools by campaign type, and watch connect rates like a hawk. The practices are well-established. The problem is the timing.

Answer rate is a lagging indicator. By the time it tells you something is wrong with a specific number, that number has already been degrading for days — sometimes weeks. The calls have gone out. The agents or AI pipelines have burned through the attempts. The damage to campaign performance has already accumulated.

Most teams discover number health issues at the aggregate level: the campaign connect rate drops, agents flag it, or a recipient mentions "your number shows as spam." At that point, the investigation begins — was it the list? The script? The timing? The agent? The number is often the last thing checked, and the diagnosis is reactive by design.

The structural gap is data visibility at the right layer

Orchestration platforms scope their analytics to the AI agent and session layer — they do not aggregate performance by individual DID across campaigns. Standard telephony Call Detail Record (CDR) reports show per-call data, not per-DID trends over time. The result: teams with rigorous number management practices are still working with a delayed, blended signal.

Vobiz Number Health surfaces per-DID analytics — answer rate, failed calls, total calls, average duration, and daily volume trends — directly in the number management dashboard, updated continuously. This guide explains how to use it, what to look for, and how to automate your response with the Vobiz API.

Understanding the degradation lifecycle helps teams intervene earlier. Most number health problems follow a predictable arc — and the window for low-cost intervention is in Stages 2 and 3, before the number reaches Stage 4.

The degradation arc — intervene in Stages 2–3, before Stage 4.

Stage 1: Healthy

The number is freshly provisioned or well-maintained. Calls connect at expected rates. Average call duration is meaningful on outbound campaigns. Carrier analytics engines have no flag on the number.

Stage 2: Early degradation

Answer rate is declining but not yet alarming. This is the stage that blended campaign dashboards most often obscure — the aggregate metric looks acceptable, but one specific DID is already trending down. Per-DID visibility catches what campaign-level aggregates miss.

Common contributing factors at this stage

• Daily call volume consistently near or above the recommended limit for that market
• A growing share of answered calls ending in short durations — a behavioral signal that carrier analytics engines associate with spam patterns
• Cold calling without opt-in context, which drives hang-up behavior and rejection patterns that degrade number reputation

Stage 3: Active degradation

The number is flagged or on the edge of flagging. Carrier analytics engines — and in India, TrueCaller — have started classifying the number as suspicious. Recipients are seeing "Possible Spam" or similar labels. Even contacts who would normally answer are declining.

Visible signals

• Answer rate in steep decline week over week
• High failure rates (CALL_REJECTED, NO_ANSWER dominant in CDRs)
• Short-duration calls even when answered — recipients hanging up immediately

Stage 4: Effectively dead

The number has been spam-labeled. Continuing to use it for outbound campaigns wastes call volume. The number needs structured spam recovery or should be released and replaced.

Number Health data surfaces five core signals. Here is what each one tells you and what to do when it moves in the wrong direction.

Signal 1 — Answer rate

What it is

Calls answered as a percentage of total calls placed on this DID.

What to watch for

A sustained decline week over week on a specific number — especially when that decline is isolated to one DID while others in the same campaign hold steady.

What to check when it drops

• If the decline is isolated to one DID, the issue is most likely that number — look at volume, duration, and failure patterns on that DID specifically
• If answer rate drops across all DIDs simultaneously, the issue is more likely the campaign — contact list quality, calling hours, or agent behavior (a prompt change, model update, or TTS voice change).

Signal 2 — Failed call count

What it is

Calls that did not connect — including NO_ANSWER, CALL_REJECTED, USER_BUSY, and ORIGINATOR_CANCEL.

Why it matters

Rising failed calls, especially CALL_REJECTED (explicit rejection by the recipient), is a strong spam signal. When a number is marked spam, recipients don't just stop answering — they actively decline. A CALL_REJECTED rate climbing as a proportion of total failed calls warrants immediate attention.

Signal 3 — Average call duration

What it is

Mean session length for answered calls.

What to look for

A drop in average duration on answered calls — especially a shift toward very short durations — indicates that people are answering and hanging up immediately. This is a trust signal, not a connectivity issue. The number's caller ID display is creating distrust before the agent can engage.

As the Vobiz Number Utilization Guide notes, short-duration calls and hang-ups are behavioral patterns that carrier analytics engines directly associate with spam classification.

Signal 4 — Daily call volume trend

What it is

The 30-day daily chart showing total, answered, and failed calls by day.

What to look for

• Diverging lines: Total calls flat or rising while answered calls trend down — early flagging signal
• Spike then drop: A heavy single-day campaign followed by sustained answer rate decline — volume threshold likely crossed
• Sudden failed call spike: May indicate carrier-level blocking or routing change on that specific number

Signal 5 — Total call volume (cumulative)

What it is

Total calls made through this DID over the 30-day window.

Why it matters

High cumulative volume on a single number significantly increases spam flagging risk, even if individual daily volumes are within guidelines. This is the signal for proactive rotation — don't wait for the answer rate to drop before moving a high-volume number to cool-down.

Number health starts at provisioning. The number series you choose directly affects pickup rates and compliance posture before a single call is made.

Number series guide for India

India's telecom regulator (TRAI) mandates specific number series for specific use cases. Using the wrong series is not just a best practice violation — it is a compliance issue that can result in campaign blocking and carrier action, both of which accelerate number degradation.

Global DIDs

For non-India campaigns, Vobiz provides DIDs in 130+ countries with outbound connectivity in 190+ countries. The same health principles apply globally — answer rate monitoring, volume guidelines, and rotation strategy — though spam flagging mechanics and regulatory requirements vary by country and carrier.

The Vobiz Number Health Analytics dashboard provides a per-DID analytics view for every phone number in your account — giving teams the per-number visibility that campaign-level dashboards in orchestration platforms and standard CDR reporting do not provide.

Who benefits from per-DID visibility

• Outbound sales and contact center teams running multiple DIDs across campaigns can immediately isolate which number is underperforming rather than investigating across the entire campaign
• In-house and self-hosted voice AI operators with no orchestration-layer analytics get their first native per-DID view without building custom instrumentation
• Voice AI operators on platforms with agent-scoped analytics (assistantId, not by phone number) gain the DID-level signal the orchestration layer does not surface
• High-volume operations running number pools across markets can track cumulative DID load across 30 days and act on proactive rotation before answer rates decline

Dashboard overview

Navigate to Phone Numbers in the Vobiz console. Select any number to open its Number Health panel. You will see:

• Number Details — status, monthly fee, purchase date, billing dates
• Number Health (Last 30 Days) — the core analytics view: Total Calls / Answered / Failed (count), Answer Rate (%), Avg Duration (m s), Total Minutes
• Daily Call Volume Chart — 30-day line chart with total, answered, and failed call lines by day

Reading the dashboard: a practical walkthrough

The figures below are illustrative examples to demonstrate how to interpret the dashboard — not benchmarks.

What to check weekly

• Answer rate per number, not just campaign average
• Failed call count trend — is it growing faster than total call volume?
• Average duration — any sudden drops indicate caller ID trust issues
• Cumulative volume — high-volume numbers need proactive rotation before answer rate degrades

Call volume is the primary controllable lever for number health. The right limits vary by market.

India: cold and warm campaign limits

The Vobiz Number Utilization Guide establishes the following thresholds for India-based campaigns.

Cold outbound: maximum 500–600 calls per number per day

Above this threshold, carrier analytics engines start classifying calling patterns as suspicious regardless of intent. This applies even when every call is legitimate and consented.

• Distribution: Spread calls evenly across the 9:00 AM to 9:00 PM window. Concentrated bursts — a large number of calls compressed into a short window — trigger pattern recognition flags even when the daily total is within limits.
• Ramp-up: Never spike a newly provisioned number from zero to maximum volume on day one. Gradual ramp-up over several days signals organic, legitimate usage to carrier systems.

Warm leads: 1,000–10,000 calls per number per day — but only when:

• Leads have existing relationships or verified prior interaction with your brand
• Average call duration remains meaningful on answered calls
• Opt-in consent is documented and traceable
• Traffic is distributed evenly across the calling window

Connectivity is significantly higher when calls are warm leads with existing relationships — and call duration and natural conversation patterns signal legitimate usage to carrier systems.

US market: volume and rotation context

For teams running US outbound campaigns, cap usage at roughly 150–200 calls per number per day, with pool rotation on a 24-hour dormancy cycle between pools — using Pool A one day and Pool B the next to allow carrier velocity checks to reset.

Monitor volume, answer rate, and duration together in Number Health. Volume alone does not tell the full story.

Rotation is the most common defense against number degradation — and the most commonly applied reactively. Proactive rotation keeps numbers alive.

The four-step rotation framework

Building a healthy number pool

Always maintain more than one active number per campaign. For high-volume outbound, segment pools strictly by use case — never using inbound support lines or warm-lead callback numbers for cold outreach. If a cold pool gets burned, your existing customer-facing numbers must remain clean.

The following steps are drawn directly from the Vobiz Number Utilization Guide, Section 5: Spam Recovery Best Practices.

When Number Health shows a declining answer rate and the issue has been isolated to a specific DID — not a list, timing, or agent issue — the following structured recovery process applies.

The Vobiz dashboard's Number Health section provides real-time insights into the performance of every DID. This data is exposed through a dedicated API, allowing you to build custom dashboards, automate monitoring, proactively rotate numbers, and detect performance degradation before it impacts your campaigns.

1 — Get Number Health for a DID

This endpoint powers the Number Health panel. It returns the overall health status of the number, summary metrics, and the data used to render the usage graph.

curl -X GET \
  "https://api.vobiz.ai/api/v1/account/{accountID}/numbers/+91xxxxxxxxxx/health?granularity=daily&days=30" \
  --header 'X-Auth-ID: {auth_id}' \
  --header 'X-Auth-Token: {auth_token}'

Query parameters

Each metric in the Number Health panel maps directly to a field in the API response:

The health rating is recalculated nightly based on the previous 7 days of activity:

2 — List all numbers

This endpoint returns all active DIDs associated with an account. Each number includes usage_status and is_spam fields, allowing you to quickly identify numbers that may require attention before fetching detailed health analytics.

3 — Understanding failures

The Number Health endpoint provides aggregated statistics. If you need a deeper breakdown of why calls are failing, use the CDR endpoint to analyze individual call records and group them by hangup cause.

curl -X GET \
  "https://api.vobiz.ai/api/v1/Account/{auth_id}/Call/?from_number=+91xxxxxxxxxx&call_direction=outbound" \
  --header 'X-Auth-ID: {auth_id}' \
  --header 'X-Auth-Token: {auth_token}'

4 — Managing number rotation

curl -X POST \
  "https://api.vobiz.ai/api/v1/account/{accountID}/numbers/purchase" \
  --header 'X-Auth-ID: {auth_id}' \
  --header 'X-Auth-Token: {auth_token}' \
  -H 'Content-Type: application/json' \
  -d '{ "country": "IN", "area_code": "80", "quantity": 1 }'

curl -X DELETE \
  "https://api.vobiz.ai/api/v1/account/{accountID}/numbers/+91xxxxxxxxxx" \
  --header 'X-Auth-ID: {auth_id}' \
  --header 'X-Auth-Token: {auth_token}'

5 — Automated number monitoring workflow

A typical system follows this flow:

Endpoint summary

Regulatory compliance is both a legal obligation and a direct input to number health. Non-compliant calling patterns do not just create legal exposure — they trigger carrier filtering, accelerate spam labeling, and degrade number reputation faster than volume issues alone.

Regulatory compliance reference table

A note on STIR/SHAKEN

STIR/SHAKEN authenticates the identity of the caller — it does not protect against spam labels from analytics engines. A fully A-attested call can still be labeled "Spam Likely" if historical call behavior or complaint rates are negative. Authentication and reputation management are separate disciplines.

DLT registration for India (1400 / 1600 series)

Using 1400 or 1600 series numbers requires completed DLT registration before any campaign traffic is sent. See the Vobiz DLT Registration Guide and 140/160 Number Acquisition Guide for step-by-step setup.

TrueCaller, CNAP, and number health in India

TrueCaller's database operates independently of carrier-level spam flags — a number can be labeled as spam by TrueCaller while still operationally clean at the carrier level, and vice versa. For India outbound campaigns, two TrueCaller practices directly affect number health:

• Register business numbers on TrueCaller Business to ensure correct business name display and prevent false spam labeling
• Include a TrueCaller identity refresh in your spam recovery process — when a number has been flagged, updating the business name in TrueCaller's database is a required recovery step, distinct from anything at the carrier level

In India, TRAI has also mandated CNAP (Calling Name Presentation) — a carrier-level standard that displays the registered business name on incoming calls. Unlike TrueCaller, which operates as an independent database, CNAP works at the network layer. For outbound campaigns, ensuring your business name is correctly registered for CNAP display is a separate step from TrueCaller Business registration, and both matter for how your number appears to recipients.

Use this checklist at campaign launch, weekly, and monthly to maintain a healthy number estate.