Forex & Brokerage

How Forex Brokerages Can Scale Outbound Calls

Teodor AvadaniTeodor Avadani, Founder·
·11 min read·Last updated:
Cover Image for How Forex Brokerages Can Scale Outbound Calls

Say your six retention agents make 300 attempts a day between them, and the dormant book sits at 30,000 accounts. To scale outbound calls, a forex brokerage has to clear three limits at the same time: how many lines can dial, how many people can take a transfer, and how many hours each country lets you call. Miss one and the other two don't matter.

This guide works the capacity math on a 20,000-account list, shows where the compliance rules cap the volume, and prices the whole run at Topcalls' $0.35 per minute. It's written for the ops or retention lead who has a deadline and a list, and no appetite for another dialer contract.

Key Takeaways

  • Three caps decide how fast a brokerage scales outbound calls: concurrent lines, agents free to take transfers, and permitted hours per country. The lowest one wins.
  • A 20,000-account list at 3 attempts each is 60,000 attempts; at 25 concurrent lines and 8 calling hours a day the campaign finishes in about 9 calling days.
  • Topcalls bills $0.35 per minute all-inclusive, so that 60,000-attempt campaign at 1.375 weighted minutes per attempt costs about $28,900, or roughly $1.44 per dialable account.
  • US rules in 47 CFR 64.1200 limit telephone solicitations to 8 a.m. to 9 p.m. local time and cap abandoned calls at 3% of live answers per 30-day campaign.
  • FCA-regulated firms keep call recordings for five years under SYSC 10A.1.14R, up to seven when the FCA asks, so storage is part of the capacity plan.
  • Topcalls handles 63,000+ AI calls a day in 32 languages, and quotes a 3-10x call-volume lift for human teams that add AI calling.

1. What does it take to scale outbound calls at a forex brokerage?

Scaling outbound calls at a forex brokerage takes three things: enough concurrent lines to cover the dialable list before the deadline, enough people to take the calls that turn into deposits, and a calling schedule that fits every jurisdiction's permitted hours. The line count is the easy part. Where most scale-up plans break is the human desk and the compliance calendar.

Start from the dialable list, not the export. A 30,000-account export from the MT4 or MT5 manager loses do-not-call flags, withdrawn consent, closed accounts, open chargebacks and numbers that fail validation before anyone dials. Call the survivors A. Every other number in the plan is arithmetic on A.

The second input is how long a line stays busy per attempt, and most attempts don't connect. If a quarter of attempts reach a person, a connected call runs four minutes and a no-answer burns 30 seconds, the weighted figure is (0.25 x 4) + (0.75 x 0.5) = 1.375 minutes per attempt. Write it down. That one number drives both the line count and the bill.

A customer reactivation campaign on Topcalls keeps the list, the retry rules and the calling windows per country inside one smart campaign, so the plan on paper and the campaign in the platform are the same object. Change the window for the Dubai segment and the throughput estimate changes with it.

2. Why does a human dialing team stop scaling?

A human dialing team stops scaling because every extra 1,000 attempts a day needs another hire, another seat licence and often another language on the floor. Agents dial in one language, work one shift, and take their connect rate with them when they leave. Topcalls quotes a 3-10x call-volume lift for human teams that add AI calling, which is roughly the size of the gap.

Picture a CFD broker with a UK licence and a Dubai office. The dormant book has English, Arabic, Polish and Portuguese speakers. Covering that with people means four hiring pipelines and four QA processes. With Topcalls it's one campaign that picks the voice per market from 32 languages.

Then there's the clock. Retail traders pick up in the evening, after the London close and before New York winds down, and the hours you're allowed to call in each country don't line up. A floor that works nine to six in one time zone misses most of the answerable window, and paying shift premiums to cover it eats the margin the campaign was meant to produce.

AttributeHire more agentsPredictive dialer + agentsAI voice agents (Topcalls)
Time to add 1,000 attempts a dayWeeks of hiring and trainingDays, if agents are freeMinutes: raise the concurrency
Languages coveredOne per agentOne per agent32 from one campaign
Cost basisSalary plus seatSeat, telephony and salary$0.35 per minute all-in
Evening and multi-zone windowsShift premiumsShift premiumsSame rate, any window
Abandoned-call exposureNonePacing against the 3% capEach line is its own agent
Three ways a brokerage adds outbound capacity
Brokerage operations lead reviewing an outbound calling campaign dashboard across time zones

The middle column is where most brokerages sit today, and the AI vs predictive dialer comparison explains why the dialer's pacing rules are the ceiling. A dialer multiplies attempts, not conversations. When it dials faster than the agents can answer, the surplus becomes abandoned calls, and the regulator counts those.

Before you pick a line count, the Outbound Calling Capacity Planning Worksheet takes you through list size, calling windows, attempts and handle time, concurrency, days to complete, human desk capacity and the minutes budget, and ends with a one-page plan you can defend in the deadline meeting.

3. How many concurrent lines does a brokerage need?

A brokerage needs enough concurrent lines to finish the list before the deadline, and no more than the human desk can absorb. Take total attempts (dialable accounts times attempts per lead), divide by attempts per hour per line, then by calling hours per day, and you have days at a given line count. Then check that against the spacing between retries, which no amount of concurrency shortens.

Here's the worked example, with the assumptions in the open so you can swap in your own:

  • Dialable list (A): 20,000 accounts after suppression and number validation, from a 30,000-account export.
  • Attempts per lead (N): 3, spaced 3 days apart, so 60,000 attempts in total.
  • Line throughput per hour: at 1.375 weighted minutes per attempt a line could make about 43 attempts an hour; a 0.8 pacing factor for retry gaps and CRM write-backs brings that to about 35.
  • Lines (L): 25 lines make about 875 attempts an hour, or 7,000 across an 8-hour calling window.
  • Throughput days: 60,000 divided by 7,000 is 8.6, so 9 calling days, or two working weeks with a holiday in one of the markets.
  • Spacing floor: (3 minus 1) times 3 days plus 1 is 7 days. Throughput, not cadence, sets the end date here.

Double the lines to 50 and throughput days drop to 4.3, but the spacing floor stays at 7. Past that point extra lines buy nothing. The reactivation call cadence guide approaches the same trade from the other side: retry rules set a floor on campaign length before a single line is provisioned.

Concurrency also has caps the vendor's quote leaves out. Each outbound number can only make so many attempts an hour before carriers start flagging it. Your CRM accepts only so many writes a minute. And the platform cap may apply across the whole account, not per campaign. The concurrent AI calls for brokers post works through each one, and the 1,000 AI sales calls a day guide shows what the infrastructure looks like at that pace.

4. What compliance limits cap outbound call volume?

Compliance caps outbound volume in three ways: permitted calling hours per jurisdiction, abandoned-call limits on automated dialing, and recording and retention duties. In the US, 47 CFR 64.1200 restricts telephone solicitations to 8 a.m. to 9 p.m. at the called party's local time and caps abandoned calls at 3% of live answers per 30-day campaign. Every line you add has to fit inside those.

The FCC rule at 47 CFR 64.1200(a)(7) says a caller may not "abandon more than three percent of all telemarketing calls that are answered live by a person, as measured over a 30-day period for a single calling campaign." The FTC's Telemarketing Sales Rule at 16 CFR 310.4(b)(4) sets the same 3% safe harbor and requires that the phone ring "for at least fifteen (15) seconds or four (4) rings before disconnecting." Both rules exist because predictive dialers dial more numbers than they have agents. An AI voice agent doesn't have that problem: each line is its own agent, so a connected call gets answered.

UK brokers carry a recording duty on top. FCA Handbook SYSC 10A.1.6R requires a firm to "take all reasonable steps to record telephone conversations" that relate to in-scope activities, and SYSC 10A.1.14R keeps those records "for a period of five years and, where requested by the FCA, for a period of up to seven years." At 60,000 attempts, recording and storage belong in the plan from day one. Topcalls records and transcribes every call inside the $0.35 rate, on infrastructure built for TCPA, TSR, DNC and GDPR.

Split the list by country and give each group its own window. Dubai and Warsaw are three hours apart, and running their windows back to back on the same 25 lines stretches the calling day without adding concurrency. Your compliance officer signs off the hours per group before the campaign is built; the compliance guide for AI calling at forex brokers has the sign-off checklist.

Retention agent taking a warm-transfer call from an AI voice agent at a forex brokerage

5. How do you keep the human desk from becoming the bottleneck?

Keep the human desk out of the bottleneck by sizing it from the transfer rate, not the line count. Connected calls per hour times the share that ask for a person gives transfers per hour; divide by what one agent handles and you have the agents needed. At 875 attempts an hour, a 25% connect rate and a 15% transfer rate, that's about 33 transfers an hour.

Eight agents taking four transfers an hour each cover those 33. Six agents don't, and the transfer queue backs up. A trader who said yes to a deposit and then sat on hold is a lost deposit, and the minutes that produced the yes are already billed.

The fix lives in the campaign rules, not in hiring. Cap transfers per hour at what the desk can take, route the overflow to a booked callback, and drop concurrency in the hours the desk is thin. The human handoff guide for forex calls covers the warm-transfer script itself.

On the data side, every disposition writes to the CRM through Topcalls' integrations with 5,000+ tools, and the real-time analytics dashboard shows the transfer queue while the campaign runs. You throttle before the queue does it for you.

6. What does scaling outbound calls cost a brokerage?

Scaling outbound calls with Topcalls costs $0.35 per minute all-inclusive, covering the voice model, telephony, recording, transcription and analytics, with no per-seat, setup or bundle fees. The 60,000-attempt campaign above at 1.375 weighted minutes per attempt is 82,500 minutes, or about $28,900, which works out to roughly $1.44 per dialable account.

Carry the per-account figure into the budget meeting, not the per-minute one. Set it against what a reactivated trader deposits and trades over the next quarter, using your own book's numbers in the dormant trader revenue calculator. The weekly burn is visible enough to stop a campaign that hasn't produced reactivations by week two, which is the stop line the worksheet asks you to write down.

On timing, Topcalls sends a proposal within 48 hours of a strategy call, first setup takes about 15 minutes, and live campaigns typically run within about 2 weeks. Topcalls runs 63,000+ AI calls a day across its customers at 99.9% uptime, so a 7,000-attempt day for one brokerage sits well inside normal operating range. Brokers who want to see the volume side first can read the sales acceleration page.

7. When doesn't scaling outbound calls make sense?

Scaling outbound calls doesn't make sense when the dialable list is under a few hundred accounts, when the accounts belong to a named account manager, when there's no consent trail, or when the human desk can't take a single extra transfer. In each of those cases more lines produce more cost and more complaints, not more deposits.

Where it doesn't fit:

  • Small lists. Under about 300 dialable accounts, an afternoon of human calls beats a campaign build.
  • High-value desks. A trader with a six-figure balance and a personal account manager expects that manager's voice. An AI opener reads as a downgrade.
  • No consent trail. Bought lists and scraped numbers sit outside the consent rules above, and volume multiplies the exposure.
  • No human capacity. If nobody can take a transfer, the AI books callbacks that nobody makes. Fix the desk first, then add lines.
  • Disputes and complaints. Accounts under a chargeback or an open complaint stay in the suppression file, whatever the volume target says.

If the dialable list has passed a thousand accounts and the team is calling a few hundred a week, the capacity math is worth 30 minutes of your time. Fill in the worksheet with your numbers, then book a 30-minute call and we'll set a concurrency figure for your account and map the disposition flow for a pilot.

Frequently Asked Questions

Get AI calling tips in your inbox

No spam. One email per week with actionable sales automation tips.

Share this article

XLinkedIn

Summarize with AI

Ready to automate your calls?

Book a 30-min call or calculate your ROI.

Related Articles