Call Routing: Strategies, Rules, and Setup Guide

Call routing is the logic that decides where each inbound call goes. Unlike simple call forwarding (which always sends to the same destination), call routing software evaluates rules in real time and selects among multiple targets—phone numbers, SIP endpoints, or browser agents—to connect callers faster, spread load fairly, and respect buyer or team constraints. In a typical flow: number → optional IVR or queue → routing rule → target → overflow to voicemail or a backup rule.

Key takeaways

• Pick a primary routing strategy (priority, weighted, round-robin, or simultaneous) and define clear overflow paths.
• Add time windows, caller-state rules, and per-target caps to improve answer rates and protect budgets.
• Keep IVR menus short; use a queue with a firm timeout; use whisper messages to give agents context.
• Test every branch before launch; monitor answer rate, duration, and recordings; adjust weights and caps weekly.
• Use spam-score monitoring, HLR/line-type checks, blocklists, and rate limits to stop fraud and short-call storms.

What call routing is and when to use it

Call routing guides each call to the best available destination based on rules you set. It differs from simple forwarding by evaluating multiple targets and choosing dynamically—by priority, weight, round-robin order, or simultaneous ring. Routes can also depend on caller state, the time of day, or per-target caps. For a practical guide to building data-driven rules (schedules, geo, IVR, queues, and caps) see our piece on intelligent call routing.

Use call routing when:

• You need to distribute calls across a team or multiple buyers.
• You operate in multiple regions, states, or time zones.
• You must enforce per-target schedules, daily/concurrent caps, or pay-per-call rules.
• You want to shorten time-to-answer without overloading any single destination.

With CallFlow, your targets can be:

• Phone numbers (mobile, landline)
• SIP endpoints (register SIP phones or PBX extensions)
• Browser agents (agents logged into a web softphone)

Where it sits in the flow: the inbound number takes the call, an optional IVR or queue segments or buys time, then the routing rule selects a target. If no one answers or a cap is hit, overflow to the next eligible target, a queue, or voicemail.

Core call routing strategies (and how to pick one)

Priority routing

• How it works: Try Target A first; if busy/no-answer/capped/closed, try Target B, then C, and so on.
• Best when: You have tiered buyers or one “top performer” you prefer to feed first.
• Notes: Combine with daily and concurrent caps to avoid starving lower-priority targets.

Example: An insurance campaign sends calls to Buyer Alpha first (top conversion), then to Buyer Beta if Alpha is closed or capped, then to your in-house desk.

Weighted routing

• How it works: Distribute calls across multiple targets by percentage (e.g., 50/30/20), only among those eligible at that moment.
• Best when: You’re balancing volume among buyers or regions based on agreements or performance.
• Notes: Revisit weights weekly based on answer rate and call outcomes. Caps still apply. For setup details, examples, and pitfalls see the weighted call routing guide.

Example: During weekdays use 60/25/15; after 5 p.m., use a scheduled rule variant with 80/20 because one partner closes early.

Round-robin routing

• How it works: Cycle through the target list, giving each eligible target an equal turn. On no-answer or busy, overflow to the next.
• Best when: You want fairness across a team of similar agents or locations.
• Notes: Pair with a short ring timeout so you don’t stall the queue. Respect concurrent caps to avoid overloading anyone.

Example: An internal sales team of four gets one call each in sequence, with 20 seconds per leg before moving on.

Simultaneous ring

• How it works: Ring multiple eligible targets at once; connect to the first to answer.
• Best when: Speed-to-answer is critical (hot leads, emergency lines, VIP service).
• Notes: Use with care—concurrent caps prevent over-alerting. Ensure agents know to stop picking up once they hear another has answered.

Example: A local emergency repair line rings two technicians and the dispatcher at once during after-hours.

Strategy comparison

Strategy	How it works	Best for	Watch-outs
Priority	Tries targets in a fixed order	Tiered buyers; top-performer focus	Starving low-priority targets; set caps
Weighted	Splits by percentage across eligibles	Buyer mix; budget allocation	Stale weights; revisit based on performance
Round-robin	Cycles through targets evenly	Fairness across similar agents	Long ring times; set per-leg timeouts
Simultaneous ring	Rings multiple; first to answer connects	Fastest answer on hot lines	Concurrency overload; educate agents

Caller-, time-, and geo-based rules that improve outcomes

Time-of-day schedules per target

Set per-target business hours so calls only route when someone can answer. Define weekdays vs. weekends; for holidays, add closed time windows as needed. When closed, either skip the target (letting the rule try others) or route to voicemail.

Configuration tips:

• Use local time zones per target.
• Keep per-leg ring time modest (15–25 seconds) to avoid long hold times.
• For teams with split shifts, create two schedule windows per day.

Geographic/caller-state routing

Route based on the caller’s state to respect licensing, service areas, or buyer exclusivity. Geographic call routing is especially useful for field services, healthcare markets with state regulations, and pay-per-call campaigns with territory agreements.

Practical patterns:

• Assign FL/GA/NC to three different teams or buyers.
• For statewide exclusivity, send all calls from that state to the assigned target; overflow to a queue or backup if closed.
• If coverage varies by day, pair state rules with per-target schedules.

Note: Mobile number portability means the area code doesn’t always match the caller’s physical location, but state-based routing still aligns well with most service and licensing needs. If precision is essential, add a one-question IVR to confirm location (e.g., “Press 1 if you’re in Florida”).

Caps and overflow paths

Use daily caps to limit total calls per target and concurrent caps to control simultaneous calls. When a cap trips, the call should overflow to the next eligible target, a queue, or voicemail.

Good defaults:

• Daily caps on buyers to enforce budgets.
• Concurrent caps of 1–3 per agent or small desk.
• Queue fallback with a 30–60 second total timeout, then voicemail.

Out-of-hours handling

• Send directly to voicemail with a clear greeting and callback promise.
• Route to an alternate team (e.g., on-call or a third-party service).
• Offer a short IVR for triage (e.g., “1 for emergency, 2 for a callback”).

Compliance and experience notes

• Keep IVR menus short (2–4 options) and avoid deep trees.
• Use plain language in prompts; if you need to notify callers about recording, add a short IVR prompt at the start stating the call may be recorded.

IVR, queues, voicemail, and whispers: structuring flows

Simple IVR menus to segment intent

A light IVR routes callers by intent without causing drop-off. One or two layers are enough.

Examples:

• Services: “Press 1 for sales, 2 for service, 3 for billing.”
• Location: “Press 1 if you’re calling about Florida, 2 for Georgia, 3 for North Carolina.”

Best practices:

• Announce the options once; keep each under five words.
• Offer a repeat option and a timeout fallback to the main queue or voicemail.

Call queue with hold music and timeout

A queue buys seconds to free up agents. Use it when you expect brief bursts of inbound volume.

Configuration:

• Queue timeout: 30–60 seconds total before overflow.
• Overflow target: next routing rule, backup team, or voicemail.
• Hold music: keep it at a comfortable volume.

Whisper messages for agent context

Whisper messages play to the agent just before connect (the caller does not hear it). Use them to tell the agent what campaign or menu path the caller came from: “Google Roofing FL lead” or “IVR 2: Existing patient.”

Two sample flows

• Lead capture hotline:
  1. Inbound number → IVR: 1=Sales, 2=Support
  2. “1=Sales” → Simultaneous ring to three agents (15 sec)
  3. If no answer → Queue (30 sec) → Priority to Team Lead → Voicemail
• Territory-based service:
  1. Inbound number → IVR: 1=FL, 2=GA, 3=NC
  2. State branch → Weighted 70/30 between two local teams (20 sec per leg)
  3. If capped/closed → Queue (45 sec) → Round-robin backup → Voicemail

Designing your routing plan: three real examples

Local service business in three states

Scenario: A roofing company runs ads in FL/GA/NC. They want the fastest answer, correct licensing, and clean after-hours handling.

Plan:

• Numbers: One toll-free on ads plus local numbers on landing pages. See our quick primer on setup in business call forwarding.
• IVR: “Press 1 for Florida, 2 for Georgia, 3 for North Carolina.”
• State branches:
  • FL → Weighted 70/30 split between Team FL-A and FL-B (Team A has higher close rate).
  • GA → Round-robin across two installers.
  • NC → Priority: in-house desk, then overflow to a trusted subcontractor.
• Schedules and caps: Each team set to its local business hours; concurrent cap = 2; daily cap tuned per crew capacity.
• After-hours: Direct to voicemail with a callback guarantee by 9 a.m. next day.
• Whisper: “Roofing Google Ads – [State]” to flag lead source and state.

Pay-per-call campaign with multiple buyers

Scenario: An affiliate network sells inbound calls from search. Three buyers accept calls with different hours and performance histories.

Plan:

• Campaign setup: Map each campaign number to a routing group with Buyers A, B, C as targets.
• Strategy: Priority routing: A → B → C based on price/performance and compliance fit.
• Buyer controls:
  • Daily caps per buyer.
  • Per-target schedules for business hours and weekends.
  • Duplicate-caller rules on the campaign to reduce re-billing disputes.
• Overflow: If all buyers are capped or closed, route to a queue (30 sec), then to a house desk (round-robin), then voicemail.
• Recording and monitoring: Enable call recording and live monitoring. If you need to notify callers, add a short IVR prompt at the start stating the call may be recorded.
• Reporting: Scheduled email reports to each buyer showing daily totals and answer rates; real-time logs available in buyer portals.
• Optimization: Re-rank priority weekly; if Buyer C’s answer rate improves, move them up or add a small weighted share in peak hours. For Google Ads traffic, confirm call paths align with landing pages—see call forwarding for Google Ads.

Multi-location medical practice

Scenario: A practice has two sites with separate front desks. They need to triage new vs. existing patients and avoid missed calls.

Plan:

• IVR: “Press 1 if you’re a new patient, 2 if you’re an existing patient.”
• Branch: New patients → Simultaneous ring to both front desks (15 sec).
• Overflow: If unanswered, route to round-robin among back-office agents (20 sec per leg), then voicemail with a same-day callback promise for new patients.
• Existing patients → Geographic routing by caller state (optional if one site handles out-of-state), else direct to site-specific round-robin.
• Schedules: Each site uses its own time windows, including configured holiday closures.
• Whisper: “New patient – Site A” or “Existing – Site B” to reduce transfer time.

Setting up call routing in CallFlow (step-by-step)

Numbers and targets

1. Provision numbers: In CallFlow, pick a US local or toll-free number (or an international number where available) from your prepaid wallet. No subscription or contract—usage-based from the wallet balance.
2. Add targets:
   • Phone numbers (enter E.164 format)
   • SIP endpoints (register SIP phones or PBX extensions)
   • Browser agents (agents log into the web softphone)
3. For each target, set:
   • Time windows (business hours, weekends, and add closed windows for holidays)
   • Concurrent cap (how many calls at once)
   • Daily cap (optional, for buyers or limited teams)
   • Whisper message (campaign/source/context)

Build rules

1. Choose a routing strategy per flow:
   • Priority: define target order and per-leg ring time.
   • Weighted: assign percentages that sum to 100 across eligible targets.
   • Round-robin: set the target list and per-leg ring time.
   • Simultaneous: choose eligible targets and a total ring timeout.
2. Add optional IVR:
   • Keep options minimal (2–4).
   • Set timeout and repeat behavior.
3. Add a queue:
   • Queue total timeout (30–60 sec is typical).
   • Hold music.
   • Overflow to next rule or voicemail.
4. Voicemail fallback:
   • Record a short, clear greeting.
   • Include any required notice language in your prompt.
5. Recording and monitoring:
   • Enable call recording if your campaign or buyer requires it.
   • Use live call monitoring during launch to spot issues early.

Test and launch

• Place test calls on every branch (each IVR option, each state rule, capped and uncapped targets).
• Confirm logs show correct routes; verify recordings; check that buyers/agents receive calls and see the right whispers.
• Validate time windows by temporarily shifting a schedule to force the closed-state behavior.
• Review analytics after day one: answer rate, ring times, and spill to voicemail.

Common mistakes to avoid

• Circular overflows: Don’t send a queue back into itself via another rule.
• Overlapping schedules: Ensure targets don’t simultaneously appear “open” across conflicting time windows.
• Missing holidays: Add closed time windows for holidays so your “closed” behavior is predictable.
• Too-short queue timeouts: If set under ~20 seconds, you may not give the next rule time to kick in.
• Unbounded simultaneous ring: Use concurrent caps so one hot lead doesn’t alert too many devices.

Measure, optimize, and scale

Analytics and reporting

• Use real-time call logs to track answer rates, durations, IVR choices, and final outcomes.
• Listen to recordings to gauge call quality and confirm buyer or agent behavior.
• Set scheduled email reports for stakeholders who need summaries without logging in.

Tuning weights and targets

• Adjust weights or target order weekly based on:
  • Answer rate (raise share to high-answer targets)
  • Handle time and close rate (favor efficient, quality outcomes)
  • Capacity signals (tighten concurrent caps if agents are overloaded)
• For simultaneous ring, periodically prune targets that rarely answer—they still consume alerts.
• When scaling to new markets, clone a proven flow and update numbers, geo rules, schedules, and whispers.

Quality control for agencies and pay-per-call

• Review buyer performance and reallocate traffic across priority or weights accordingly.
• Watch duplicate-caller rules behavior; align with buyer expectations to minimize disputes.
• Confirm IVR menus still match landing-page promises and ad copy to reduce early hang-ups.
• For toll-free campaigns, keep an eye on number reputation and spam flags—see our primer on toll-free number spam score.

Troubleshooting, spam protection, and safeguards

Spotting spam/fraud and short-call storms

• Monitor per-number spam scores; a sudden drop in answer rate or a spike in very short calls often signals a reputation issue.
• Use HLR/line-type lookup to identify risky routes; block or rate-limit known VoIP ranges that flood your numbers.
• Maintain blocklists for abusive caller IDs and set per-number rate limits to throttle bursts.

Protecting your numbers and routes

• Keep IVR prompts short and professional—spam dialers tend to hang up early when they hit real menus.
• Rotate or pause numbers that show persistent reputation issues; monitor improvements before reintroducing them.
• Behind the scenes, multiple redundant carriers and direct SIP trunks increase resilience; still, design overflow paths you control.

When calls don’t connect

• Check call logs for call status/reason codes (e.g., busy, no-answer, rejected). For details on interpreting SIP responses and related issues, see: SIP response codes and spam detection.
• Validate target status: wrong number, SIP registration expired, or browser agent logged out.
• Confirm schedules and caps; a “closed” target won’t ring even if it sits earlier in a rule.
• Build a test plan after any change; keep a rollback path ready if KPIs dip.

Create an account at /register or talk to a specialist at /contact.