Call RoutingNovember 18, 20269 min read

Agency resell call routing: setup with sub-accounts and client access

Blueprint for agencies to provision numbers, isolate client data with sub-accounts, expose logs and reports, and keep carrier plumbing centralized.

CallFlow Engineering Team

Telecom engineering and support at CallFlow (ALERTSIO LLC)

Agency resell call routing: setup with sub-accounts and client access – Call Routing article from the CallFlow blog

Use a routing platform with sub-accounts, a central prepaid wallet, and buyer or client portals to run agency resell call routing. You provision numbers, assign them to each client’s workspace, and keep the carrier routes and billing centralized while clients see only their own calls and recordings. By the end you’ll have one client live with numbers, routing, access, and safeguards, ready to clone to other clients. Expect about an hour for the first build and much less for copies.

The short answer: agency resell call routing and numbers

You run one master account with a prepaid wallet and carrier-grade routes, then create a sub-account (a separate client workspace) for each client or buyer. You buy numbers centrally, attach each number to that client’s routing, and grant that client a login with view or limited-edit access. All carrier connections, spend, and controls stay in your master account. Clients get their own logs, recordings, analytics, and scheduled reports, and cannot see anything from other clients.

Account structure that keeps clients isolated

Sub-accounts are separate client workspaces under your agency umbrella. Your agency admin owns the wallet, numbers, routing designs, and global settings. Client or buyer logins sit only in their sub-account. You can set them to view-only or allow limited edits you’re comfortable with (for example, business hours, destination numbers, or greeting uploads).

Map a simple setup. You manage three clients. Each client has its own campaigns, numbers, targets (the phones or SIP endpoints you forward to), and call logs. Client A never sees Client B’s calls or recordings. You can still copy a routing design from Client A and reuse it for Client C without exposing data.

Who should own numbers? Keep numbers under the agency master so you can manage billing, swapping, and QA (quality checks) without waiting on client credentials. If a client asks to keep a number, spell out ownership and transfer rules in your contract. If a move is required, the destination provider’s port-in process and the country’s rules will apply. For coordination questions, reach us at Contact.

Data isolation is the point of this model. Clients and buyers see only their own live logs, recordings, and reports. Your team can work from India, the Philippines, Thailand, Bangladesh, or Pakistan with the same structure. Routing stays centralized, and client access stays scoped.

As a quick aside: on CallFlow you create sub-accounts for clients and buyer logins inside them, with all routing and carriers managed in one place.

Provision numbers for clients from one wallet

You buy and assign numbers from your master wallet, then move them into each client’s workspace. This keeps purchasing and carrier plumbing under your control while giving clients their own numbers and logs.

For US callers, pick toll-free when national reach and brand recall matter. Use local numbers if the buyer wants a local presence. If your client needs several toll-free lines for tracking and overflow, see managing multiple toll-free numbers.

If your agents or buyers answer outside the US, route to direct SIP endpoints (a VoIP address) or mobile numbers as the client prefers. SIP is often better for call quality and for handling more than one call at a time.

Name hygiene helps later audits. Label numbers by campaign and source (for example, “ClientA-Google-Search-West”). Add a short whisper (a brief audio the target hears before connect) so the answering agent knows the campaign.

Procedure:

  1. Choose the country and area code, then pick the quantity. For US campaigns, decide on local area codes by market or go with a toll-free if reach matters more than locality.
  2. Assign each number to the right sub-account so it inherits that client’s routing and logs. This keeps client data separate while you still own and manage the numbers.
  3. Add tags and a clear label per number. Use a consistent pattern you can search later, like “Client-Channel-Region.”
  4. Set the initial destination: a SIP URI, a buyer’s phone, or a softphone agent. If you plan an IVR (an automated menu) or queue, point the number to that entry point instead.
  5. Add a whisper and call recording if the client allows it. Enable call recording at the campaign level so later routing changes don’t affect your compliance approach.
  6. Save and place a quick test call to confirm the number rings somewhere under the client account. You’ll refine routing next.

Compliance note: turn on call recording only where allowed. Follow the call recording laws for your country and, for US work, the caller and receiver states. If in doubt, announce recording up front.

Build the routing your clients expect

Weighted routing with overflowDiagram of Weighted routing with overflow. Steps: Caller; 888-201-4000 (Toll-free); IVR (Press 1 Sales); Router (Weighted 50/30/20); Buyer A (9–17 local); Buyer B (9–17 local); Buyer C (10–18 local); Queue (Hold 30s); Voicemail. Connections: Caller to 888-201-4000; 888-201-4000 to IVR; IVR to Router (Press 1); Router to Buyer A (50%); Router to Buyer B (30%); Router to Buyer C (20%); Buyer A to Queue (no answer); Buyer B to Queue (after hours); Buyer C to Queue (no answer); Queue to Voicemail (30s).Press 150%30%20%no answerafter hoursno answer30sCaller888-201-4000Toll-freeIVRPress 1 SalesRouterWeighted 50/30/20Buyer A9–17 localBuyer B9–17 localBuyer C10–18 localQueueHold 30sVoicemail
Illustrative exampleWeighted routing with overflow Sales calls split 50/30/20, then overflow to queue and voicemailSwipe sideways to see the whole diagram.

Most client needs fall into a few patterns you can reuse:

  • Priority fallback: ring Buyer A first, then B, then C if no answer.
  • Weighted split: send a share of calls to each buyer (for example 50/30/20), good for pay-per-call.
  • Round-robin: cycle through buyers to keep load even.
  • Simultaneous ring: ring several targets at once, connect the first that answers.
  • Time-of-day schedules: route based on open hours and after-hours rules.

Build a short IVR to sort intent if needed. One or two options are enough: Press 1 for Sales, 2 for Support. For Sales, drop into the router. Keep ring timers tight per target, usually 15–25 seconds, so you don’t hit the buyer’s personal voicemail before your overflow. If a buyer uses a mobile, shorter is safer.

Geo rules help when buyers own territories. For US work, route by caller state or by area code so California goes to the West buyer and New York to the East buyer. For other countries, use the best available caller region data.

Carry one example through your first client. The toll-free number 888-201-4000 points to an IVR. Press 1 for Sales to enter the router. Use a weighted split of 50% to Buyer A, 30% to Buyer B, and 20% to Buyer C. Set Buyer B to pause after 17:00 in their local hours and on weekends. If no one answers within 20 seconds, send the caller to a short queue (hold music) for up to 30 seconds, then go to voicemail. After hours, skip buyers and go straight to the queue, then voicemail.

Per-target call windows matter. Buyers want to sleep, and you don’t want a buyer cell to grab the call with personal voicemail. Use strict open and close windows per buyer and a clean after-hours path.

If you run pay-per-call programs, you’ll likely pair weighted splits with buyer portals and caps. A refresh on the core building blocks is here: pay-per-call setup.

Give clients access to call logs and recordings

Turn on client or buyer portal access so each party can see their own activity in real time. They should see caller ID, call start time, and duration. They should also see the route the call took (IVR and routing path), the recording link when enabled, and simple analytics like answer rate and time to answer. Point them to scheduled email reports so they don’t need to pull data every day.

Scope access by who’s logging in. Client-wide access lets a client see all calls and recordings for their numbers and campaigns. Buyer-only access limits visibility to the calls that buyer received, not calls sent to others. You can keep access view-only, or allow limited edits to safe fields like open hours, destination numbers, and voicemail greetings. Keep routing weights and geo rules under agency control.

Operational habits reduce ticket churn. Send a weekly performance report and a daily missed-call digest to prompt follow-ups. For deeper drill-downs and trends you can reference or share selectively, see call analytics basics.

Safeguards: caps, duplicates, spam and fraud controls

Caps (limits on daily or concurrent calls) keep your budget and buyer promises intact. Use daily and concurrent caps per buyer so one buyer doesn’t take the entire day’s volume. Define what happens at the cap: roll to the next buyer in the router, push to a queue, or drop to voicemail. Set caps on the campaign too when you must protect top-line spend. If you’re new to caps, this walkthrough helps: daily and concurrent call caps.

Duplicate-caller windows prevent repeat billing and a poor caller experience. If the same caller returns within your window (for example, a few hours), send them to the buyer that took the first call. Or route them to an alternate path that doesn’t bill a second time. The goal is fair buyer credit and less friction for the caller. More on setups here: duplicate call suppression.

Spam and fraud controls save time and money. Watch the per-number spam score and rotate a number out if it starts showing “Spam Likely” to callers. Maintain blocklists so you can stop known bad callers or sources at the number or account level. Consider VoIP-caller blocking (screen internet-originated calls) if your campaigns attract bot traffic. Use rate limits (control how many calls you accept in a burst) when floods hit to protect buyers.

Use live call monitoring (watch calls as they arrive) to spot short-call spikes and bad answer behavior. A wave of 0–6 second calls points to spam or filtering upstream. Repeated no-answers point to poor buyer pickup or a mobile voicemail that grabs calls too fast. Adjust routing, caps, or blocks in response.

Operate and hand off: testing, QA, and reporting routines

Treat launch as a checklist. Make a few live test calls for each IVR option. Let each buyer ring long enough to verify the timer, then watch the overflow hit the queue and voicemail. Confirm the whisper plays and the recording, if enabled, attaches to the call log. Check the buyer’s phone actually rings (their SIP phone is registered, and their mobile has coverage).

If you use geo or state routing, map test numbers and place calls that originate from each region (use numbers with those area codes) to verify the path. Place a test, watch the live log, and confirm the call path and buyer match the rules.

Set agency defaults and stick to them unless the client requires a change:

  • Ring time per target: 20 seconds is a safe start to outrun mobile voicemail.
  • Queue length: short holds reduce abandon rates; 30 seconds is common.
  • After-hours: straight to voicemail or a short queue, then voicemail, with a clear greeting.

Ongoing quality checks keep results steady. Review answer rate, time to answer, short call rate, and number spam score each month. Shift router weights toward buyers who answer fast and convert, and away from those who miss calls or let voicemail pick up. Update open hours for holidays and seasonal changes in each buyer’s local hours.

Hand the client a tight packet so they can self-serve. Include the portal link, their login scope (client-wide or buyer-only), what they can see and change, the reporting schedule and recipients, and who to contact for routing or number changes.

How to confirm it worked:

  1. Place three test calls: during open hours, after hours, and when a buyer is at cap.
  2. Watch the live log show the IVR choice, the target rung, and the overflow path.
  3. Log in as the client or a buyer and make sure they see only their calls and recordings.
  4. Check scheduled reports arrive with the right data the next morning.

Next step: build one client end to end, then clone the pattern to others. If you want to manage this with a central wallet, multiple redundant carriers behind the scenes, sub-accounts, and buyer portals, you can create a workspace at our signup page.

Frequently asked questions

Can I white‑label the client portal for my agency brand?

You can create sub-accounts and buyer portals that show only each client’s numbers, calls, recordings, and reports, keeping your agency controls and carrier plumbing hidden. The platform documentation does not promise a built-in full white‑label rebrand for every element. If you need custom branding or a branded login experience, contact support to discuss whether the portal can be adapted for your agency.

Do clients need special software to access call logs and recordings?

No special desktop software is required. Clients access real‑time call logs, recordings, and scheduled reports through their portal in a web browser. If they need to answer calls, they can use a browser softphone, register a SIP softphone, or receive calls on any phone number you forward to. Recording links play in the portal when call recording is enabled and allowed by law.

Can I route to a buyer’s SIP phone or a browser agent instead of a mobile number?

Yes. You can forward calls to direct SIP endpoints (SIP URIs) or to browser agents (the browser softphone), as well as to regular phone numbers. Use SIP or the browser softphone when buyers need better audio quality or need to handle multiple concurrent calls. Set per‑target schedules and caps so calls only go to a buyer’s SIP or browser agent when they are available.

Will US toll‑free or local numbers still work if our agents answer in India or the Philippines?

Yes. US toll‑free and local numbers can forward to agents answering from India, the Philippines, or other supported countries. Route the number to a SIP endpoint or to the agent’s phone. Remember to configure per‑target windows and ring timers so personal voicemails aren’t captured. Enable call recording only where it is legally allowed for the caller and receiver locations.

Who should own the numbers, and can they be ported if a client leaves?

For operational control, keep numbers in the agency master wallet so you can manage purchasing, swaps, and QA without needing client credentials. Spell out ownership and transfer terms in your contract. If a client insists on keeping a number, follow the destination provider’s port‑in process and the applicable national rules; CallFlow support can help coordinate, but the final port depends on the receiving provider and country regulations.

How do I handle time zones when clients and buyers are in different countries?

Use per‑target schedule windows (time‑of‑day) and buyers' local hours so each buyer only receives calls during their workday. Combine schedules with geo rules (caller state or area code) when territories matter. Set ring timers to 15–25 seconds to avoid mobile voicemail, and use after‑hours paths (queue or voicemail) so calls have a clean fallback when no buyer is available.

From the team

CallFlow Engineering Team

Telecom engineering and support at CallFlow (ALERTSIO LLC)

The engineers and support staff who build and operate CallFlow's call-routing platform. We write from what we see running inbound routing for pay-per-call marketers, agencies and small businesses every day: routing rules, carrier behaviour, spam flags, and the configuration mistakes that quietly cost calls.

Ready to get started with CallFlow?

No subscription. Transparent usage billing. Per-number spam scores. Built for scale.

Create Free Account →