Call Analytics: Metrics, Setup, and Real Examples

Call analytics is the discipline of measuring inbound phone performance so you can route more calls to answers, cut waste, and show results to stakeholders. In a call-forwarding environment, that means using call logs and routing events—not transcriptions—to understand who called, where the call went, what happened at each hop, and how long it took.

Key takeaways

• Good call analytics ties every call to a number, campaign, route, and target so you can take action.
• The most useful metrics are simple: answer rate, ring time, duration, queue wait, and per-target performance.
• SIP response codes and short-call patterns tell you what’s failing: busy targets, blocked calls, or upstream refusal.
• Provision numbers per channel/geo for clean attribution; use schedules, caps, and routing types to test and improve.
• Share weekly reports with 3–5 KPIs and clear to-dos; keep live logs open during spikes to validate traffic quality.

What is call analytics? (and what it isn’t)

Call analytics, in this context, is the measurement of inbound phone call data and outcomes across numbers, routes, and targets. It’s about the mechanics of how a call reached you and what happened—built from the same logs your routing uses to make decisions.

There are three useful data layers:

• Network signaling: SIP response codes, connect time, and hang-up reason. These explain why calls fail or connect (486 Busy Here, 603 Decline, 607 Unwanted), how long it rang, and who terminated the call.
• Routing events: IVR selections, queue entries/exits, which targets were tried and in what order (priority, weighted, round-robin, simultaneous), whether a schedule window was open, and whether caps were hit.
• Business context: the dialed number (and its campaign/source), caller state, duplicate-caller rules, buyer or sub-account assignment, and recording availability.

What this article does not cover: AI transcription, keyword spotting, or sentiment analysis. We’ll stay focused on routing and performance analytics derived from call logs and events—what most SMBs, agencies, and pay-per-call teams actually use to tune outcomes.

UCaaS/phone system analytics vs. routing analytics: If you run Microsoft Teams, its call analytics shows per-user call quality and device/network stats—excellent for troubleshooting a headset or Wi‑Fi issue. Marketers and call ops, however, need number/campaign-level routing analytics tied to attribution and targets. Think of Teams or other VoIP analytics as user quality dashboards; routing analytics is your inbound traffic control. For background on routing strategies, see intelligent call routing.

The call analytics metrics that actually move results

Focus on metrics you can act on. Here’s what to track, how to calculate it from standard call logs, and what each one tells you.

Core KPIs

• Total calls and unique callers: Count calls and de-duplicate by caller ID within a period. Use unique callers to normalize answer and duration trends; spikes in totals but flat uniques often mean retries.
• Answer rate: Answered calls divided by total connect attempts. If answer rate slips, look at ring time, schedule coverage, and per-target answer rate next.
• Average ring time (time to answer): Time from offering the call to a target to answer or timeout. Long ring times depress answer rate; consider simultaneous ring or more targets.
• Average and median call duration: Duration after answer. Median is harder to skew; high short-call share with low medians usually signals low intent, wrong routing, or blocked spam.
• First-time vs. repeat: Tag a call as first-time if the caller ID hasn’t appeared in your window (e.g., 30 days). First-time share helps read campaign effectiveness and buyer acceptance.
• Voicemail rate: Calls that finished in voicemail divided by total. Rising voicemail rate with normal volume suggests staffing or schedule issues.
• Queue wait time (if used): Time in queue before answer or abandonment. Watch abandonment vs. wait time to decide whether to keep a queue or move to voicemail.
• Abandonment before answer: Calls that hang up before a target answers, including during IVR, ring, or queue. Segment by where the drop happened.
• Time-of-day/day-of-week distribution: Plot volume and answer rate by hour and weekday to set schedules and caps.

Routing/target KPIs

• Per-target answer rate and average duration: For each forwarding target or SIP endpoint, track how often they answer and how long calls last. Low answer with normal ring time suggests availability; low duration with normal answer can flag misrouted or low-intent traffic.
• Connect time per route: Compare simultaneous ring vs. round-robin. Simultaneous ring typically lowers time to answer; round-robin can balance load but add ring.
• Caps hit and schedule mismatches: Count calls blocked or rerouted because daily or concurrent caps were reached or schedules were closed. Frequent hits mean you’re under-provisioned.
• Buyer acceptance rate (pay-per-call): Percent of delivered calls that meet a buyer’s operational definition (e.g., answered within their window). Use with duration distribution to flag returns risk due to short calls.

Marketing/attribution KPIs

• Calls by source/number/campaign: Each number maps to a channel (Google Ads, LSA, mailer). Provision per source per geo for clean slices.
• Qualified-call rate: Define it operationally (e.g., answered and >N seconds, or answered and reached a specific IVR path). Track by campaign to cut waste.
• Duplicate-caller share: Portion of calls flagged as dupes in your rule window. High dupes in retargeting channels are common; price your time accordingly.

Quality and failure indicators

• SIP response patterns: Watch 486 (busy), 603 (decline), 607 (unwanted), 408 (timeout). A jump in 486 means targets can’t keep up; 603 can indicate manual declines; 607 and high early-cancel can suggest spam labeling or call screeners. See our deep dive on SIP response codes.
• Short-call rates: Spikes in sub-10-second calls often mean wrong numbers, fraud, or spam filtering upstream.
• Spikes by caller state or line type (e.g., VoIP): Concentration by state or VoIP line type can indicate a campaign shift or abuse.

Common pitfalls

• Using average duration as quality truth: Duration is a proxy; segment by IVR path, target, and campaign before drawing conclusions.
• Ignoring ring time: Long ring harms answer rate; fix routing first, not targets.
• Mixing first-time and repeat: Repeat callers skew duration longer; separate the series.

How call analytics works in a call-forwarding platform

Attribution starts with numbers. Provision a unique number per campaign or channel—and often per geography—so every call inherits clean source metadata. When a call arrives, the platform logs:

• Call start and end timestamps
• Caller ID and detected state
• Dialed number (and its campaign)
• Routing steps: IVR selections, queue entry/exit, schedule decisions, caps hit
• Targets tried and in what order; which target answered (phone, SIP endpoint, browser agent)
• Final outcome: answered, voicemail, abandonment, blocked (with SIP codes)
• Durations: total, ring/queue, talk time
• Recording status and file (if enabled)

Configuration creates analyzable data. The more intentional your routing, the clearer your analytics:

• IVR menus that split service lines or prequalify
• Call queues with hold music for team handling
• Time-of-day schedules per target and per campaign
• Geographic/caller-state routing for coverage and compliance
• Priority, weighted, round-robin, or simultaneous ring to balance speed vs. fairness
• Per-target daily and concurrent caps to prevent overload and enforce buyer terms
• Voicemail with clear business hours and whisper messages to guide targets

Operational use

• Real-time call logs and analytics views let you watch answer rate by hour and fix routes on the fly.
• Scheduled email reports keep executives and buyers aligned with trend deltas and to-dos.
• Live call monitoring helps validate traffic quality during campaigns and troubleshoot edge cases.

Setup walkthroughs for three common scenarios

Local services business (multi-channel marketing)

Objective: Attribute calls by channel, route to the right on-call techs, and improve answer rate without staffing a full call center.

Steps

1. Provision numbers

• One local number per channel: Google Ads, Local Services Ads, mailers, website header.
• Add a toll-free if you advertise across regions and want a national touch.

1. Configure call flow

• IVR: “Press 1 for new service, 2 for dispatch.” Route 1 to sales/onboarding; 2 to on-call techs.
• Routing: Simultaneous ring the on-call mobile and the office SIP/desk phone; keep ring durations short before failover to the next target or voicemail.
• Fallbacks: If no answer, send to a small queue with hold music, then voicemail.
• Schedules: Business hours for sales; 24/7 for dispatch with a different ring group on nights/weekends.
• Recording: Enable for the new service line to audit lead quality and train staff.

1. Define success and reporting

• Success KPIs: Answer rate (by IVR path), average ring time, median duration, voicemail rate, and calls by channel.
• Weekly report: Top channels, answer vs. miss trend, recordings to review five edge cases.

Week 1 analytics to watch

• Answer rate by channel and hour; ring time for dispatch vs. sales.
• Short-call share by channel; if one source drives many <10s calls, adjust targeting.
• Voicemail rate after hours; ensure the dispatch line truly covers nights/weekends.

Week 2 decisions

• If answer rate lags on sales, switch from round-robin to simultaneous ring between office phones and a browser agent.
• If dispatch ring time is long, add a third on-call target or use simultaneous ring across more devices to reduce time before failover.
• Add new numbers for any new ad groups/ZIP clusters so attribution stays clean.
• Consider geographic routing if a few states dominate and you want region-specific handling.

Pay-per-call campaign splitting calls across buyers

Objective: Deliver qualified calls to multiple buyers, enforce rules, and control returns risk.

Steps

1. Create a campaign and numbers

• Provision dedicated numbers per traffic source so you can see buyer performance by source.

1. Add buyers as targets

• Each buyer is a target with their own portal access.
• Enforce duplicate-caller rules (e.g., within X days) at the campaign level.

1. Routing and controls

• Use weighted or priority routing based on buyer capacity and acceptance history.
• Set per-buyer call windows to match when they answer.
• Enforce daily and concurrent caps to protect buyer terms.
• Add whisper messages to identify the campaign/buyer to the target on answer.

1. Tracking

• Monitor per-buyer answer rate, ring time, median duration, and short-call spikes by source.

Week 1 analytics to watch

• Caps hit per buyer; closed schedule mismatches.
• Weighted vs. actual distribution; adjust weights if a buyer under-answers.
• Short-call spikes; if one source drives many sub-10s calls to a buyer, consider source-level filters or route that source to a different buyer.

Week 2 decisions

• Re-weight or re-order buyers based on answer rate and duration.
• Expand or tighten call windows to align with actual buyer staffing.
• Add geographic splits if one buyer outperforms in certain states.
• Provision additional numbers per top-performing source for A/B testing.

Inbound team with agents on schedules

Objective: Centralized inbound line with IVR and a small team that rotates coverage.

Steps

1. Provision numbers

• One local and one toll-free number. Publish locally for trust; use toll-free for national ads.

1. Configure flow

• IVR: Route common intents (sales, support, billing).
• Queue: Enable a short queue with hold music; overflow to voicemail if no answer.
• Agents and devices: Use round-robin to balance load; allow a browser or SIP softphone for remote staff.
• Schedules: Per-target or group schedules with clear open hours.

1. Monitor and iterate

• Track answer rate by hour, queue wait before answer, and missed-to-voicemail ratio.
• Listen to a sample of recordings to verify IVR intent mapping.

Week 1 analytics to watch

• Hourly heat map: volume vs. answer rate.
• Queue time vs. abandonment rate.
• Track per-target answer rate and ring time; look for outliers.

Week 2 decisions

• Shift schedules 30–60 minutes earlier/later where answer dips.
• Add a simultaneous ring overflow group during peak hours.
• If voicemail spikes at lunch, stagger breaks or add a temporary cap to slow inflow.

Finding and fixing issues using analytics

SIP codes tell you what failed and what to do next

• 486 Busy Here: Targets are ringing but not available. Add simultaneous ring targets or adjust schedules.
• 603 Decline: Manual rejects. Confirm target behavior; if expected, add overflow routing.
• 607 Unwanted: Often associated with call screening/spam. Rotate numbers, review spam-score trends, and consider caller vetting.
• 408 Request Timeout or repeated 480 Temporarily Unavailable: Target endpoints unreachable; verify SIP registrations or PSTN health.

See our SIP codes guide for details. The SIP protocol is documented by the IETF (RFC 3261) at https://www.rfc-editor.org/rfc/rfc3261.

Watch spam-score trends and short-call spikes

• Per-number spam-score monitoring and a rise in sub-10-second calls can signal labeling issues or low-intent traffic.
• Actions: Rotate or retire affected numbers, tighten duplicate rules, adjust routing to better-answering targets, and retrain IVR prompts. For toll-free behavior in particular, see toll-free number spam score.

Detect routing misconfigurations quickly

• Caps exceeded: Calls bounce or reroute early; raise caps or spread load across targets.
• Closed schedules: Answer rate drops at specific hours; confirm time zones and schedules.
• Unreachable endpoints: SIP devices not registered or PSTN targets invalid; place a quick test call and check real-time logs.
• Queues without available agents: Calls sit until timeout; either lower the queue length or ensure targets are reachable (phones not in DND, endpoints registered).

Fraud and abuse indicators

• Sudden bursts from one state or NPA-NXX.
• Many very short calls with repeating caller IDs or VoIP line types. Responses: Apply rate limits, add blocklists, enable VoIP-caller blocking, and use live monitoring to sample traffic. See live call monitoring and fraud prevention.

Quick diagnostics when answer rate drops

• Time window check: Are you open in the right time zone? Any recent schedule edits?
• Target health: Are SIP endpoints registered and PSTN numbers reachable? Any devices in do-not-disturb?
• Routing mix: Did you change from simultaneous ring to round-robin, or alter the sequence so calls spend longer on early targets before failing over?
• Caps: Are daily or concurrent caps hitting earlier due to volume shifts?
• Carrier paths: CallFlow uses multiple redundant carriers behind the scenes; if one path has issues, traffic fails over automatically.
• Recent edits: Review recent configuration changes (IVR edits, number swaps, or new block rules) made by your team.

Reporting and sharing the data

Build reports people read and use

• Weekly stakeholder report: Top numbers/campaigns, answer vs. miss trend, per-target performance, and 3–5 recordings to review edge cases. Add concise “next actions” (e.g., expand dispatch window by one hour).
• Buyer reports: Calls delivered, answer rate, qualified-call rate (your operational definition), duration distribution, and duplicate hits. Keep it consistent week over week.

Automate and stay real-time when needed

• Use scheduled email reports for executives, buyers, and partners. Include small trend deltas (“+3 pts answer rate”) and a to-do list.
• During campaign launches or spikes, keep live logs open and sample recordings hourly to validate traffic quality and routing intent.

Access and visibility

• Give buyers and sub-accounts access only to their calls, targets, and recordings. Granular access protects data while letting each party act quickly.
• Team leads can monitor per-target performance to coach and adjust routing.

Choosing call analytics tools: approaches compared

Different teams need different lenses. Here’s how common approaches line up.

Approach	Best for	Data you get	What you can change in real time	Typical blind spots
Phone system/UCaaS analytics (e.g., Microsoft Teams user/call quality views)	IT/helpdesk troubleshooting for employee calling quality	Per-user/device call quality, jitter/packet loss, endpoint stats, basic call outcomes	Device settings, network/QoS guidance, user policies	Campaign/source attribution, multi-target routing steps, buyer-level outcomes
Raw carrier CDRs	Accounting, audit, and coarse traffic summaries	Start/end times, A/B numbers, duration, high-level cause codes	Little to none from within CDRs	IVR/queue steps, per-target attempts, campaign context
Call-routing/forwarding platform analytics (like CallFlow)	Marketers, pay-per-call teams, and inbound ops who route at scale	Numbers tied to campaigns, routing events, targets tried, schedules/caps decisions, SIP codes, recordings	Change routing strategies, schedules, weights, caps, block/allow rules immediately	Deep user device/network diagnostics

When you need each approach

• If your pain is headset/Wi‑Fi quality or user adoption, UCaaS/phone system analytics is the right tool. Microsoft Teams call analytics is strong for per-user quality troubleshooting and policy-driven environments. See Microsoft’s documentation at https://learn.microsoft.com/microsoftteams/use-call-analytics-to-troubleshoot.
• If you must reconcile invoices or validate high-level volumes, raw CDRs do the job.
• If you’re optimizing campaigns, buyers, or inbound staffing, you need routing-level analytics tied to numbers and targets. That’s where call analytics software designed for call-forwarding shines.

Practical wrap-up

• Provision numbers per campaign and geo so attribution is clean from day one.
• Turn on recording where appropriate, and define “qualified” operationally so reports are actionable.
• Start with simple routing—e.g., simultaneous ring for speed—then tune per-target schedules, weights, and caps based on week-1 data.
• Keep an eye on SIP codes and short-call spikes; rotate numbers and adjust routing before problems snowball.
• Set up weekly scheduled reports with 3–5 KPIs and next actions; keep live logs open during launches.

If you route inbound calls at scale and need analytics you can act on, create an account and test a few numbers in minutes at our signup. If you want help mapping your current data to a cleaner routing plan, reach out at our contact page.