Sales and Retention Attrition: The Hiring and Retraining Cost Forex Brokers Ignore

On most broker desks, attrition is handled as an HR number and a mild irritation: somebody resigns, a replacement is found, the team moves on. The accounting view is different. Every seat that turns over costs you a month of reduced output before the exit, recruiter fees and interview hours, weeks of full pay before the new hire is allowed near a live client, supervisor hours that come out of somebody else's job, and a book of enquiries and clients that sits unattended in between. On a regulated or regulation-adjacent desk, the retraining half of that bill is larger than on a general sales floor, because a new caller cannot simply be handed a phone. This article prices one departure end to end, from an Indore back office to a Dubai desk, and turns your own turnover into an annual figure.

Why turnover costs more on this desk than on a general sales floor

On a general telesales floor, a new hire learns a product, a script and a CRM. On a broker sales or retention desk, the same hire must also learn which jurisdictions the desk may lawfully contact and on whose licence, what the approved wording is, which risk disclosures must be given and when, what may never be promised, and how the onboarding and verification steps work before an account can be funded. None of that is optional and none of it can be compressed into an afternoon.

That is not a compliance department preference. The Reserve Bank, setting out that residents may undertake foreign exchange transactions only with authorised persons and for permitted purposes and that electronic execution is confined to authorised platforms and recognised exchanges, specifically flagged "agents" who personally contact people with promises of exorbitant returns alongside misleading online advertising (Reserve Bank of India). The behaviour regulators worry about is precisely what an undertrained, under-supervised, incentive-pressured new caller drifts towards. Every unplanned departure resets somebody's training clock and widens that gap for a few weeks.

There is a second asymmetry. Acquisition callers, onboarding staff and retention staff are not interchangeable, so a resignation in one of the three cannot be covered by moving a body from another for long. Retention in particular carries relationships: when the person leaves, the client does not automatically transfer their trust to whoever picks up the file next. None of this is legal advice, and your own training obligations should be confirmed with a qualified lawyer and your regulator.

What the data says, and what drives churn on a forex desk

Start with what can be said honestly. Aon's 32nd annual salary increase and turnover survey, released in February 2026 and covering more than 1,400 organisations across 45 industries, put overall attrition in India at 16.2% in 2025, down from 17.7% in 2024 and 18.7% in 2023, and projected average salary increases of about 9.1% for 2026 (Aon). That is a cross-industry figure, not a forex-sales figure, and it should be used as context for direction rather than as your planning assumption.

What cannot be said honestly is anything specific about turnover on broker sales desks. The research behind this batch found no credible, independently sourced attrition figure for forex sales, retention or telecalling roles in India, the UAE or Sri Lanka. Vendor blogs circulate wide ranges for voice and call-centre attrition that contradict each other and cite no methodology, so none of them appears here. Your own exit register over the last twelve months is better evidence than any of them, and it takes an hour to build.

On causes, stay qualitative. Four recur in operator conversations: incentive plans that swing hard with lead flow, so a weak month in sourcing becomes a pay cut the caller did not cause; the grind of cold outreach when lead quality is poor; the pressure of working inside scripts and disclosure requirements a new hire was never properly trained in; and promotion ceilings on small desks. Three of the four are sourcing and management problems wearing an HR costume, and they form a loop: poor lead flow raises churn, churn lengthens ramp, longer ramp wastes more of the leads you do have.

The replacement cost stack for one seat

Price one departure end to end. The right-hand column tells you where your own figure lives.

Cost line	Type	Where to find your figure
Notice-period output loss	Hidden	Last month's output vs their average
Recruiter fees and job ads	Cash	Invoices for last year's hires
Interview and selection time	Allocated	Hours x cost per hour of interviewers
Onboarding and access setup	Cash	Admin hours, SIM, licences, devices
Ramp pay before live calling	Cash	Weeks of full pay before sign-off
Supervisor and reviewer time	Allocated	Coaching and call-listening hours
Orphaned leads and clients	Hidden	Records unworked during the gap

External benchmarks exist but do not transfer. An article from SHRM reports an average cost per hire of close to USD 4,700 for non-executive roles and quotes a consultant's estimate that the total cost to replace someone can reach three to four times the position's salary, split roughly between hard costs and soft costs such as manager time. That is United States data and the multiple is an expert estimate rather than a survey finding, so do not convert either into rupees and call it your number. Use the structure, not the figures.

The two hidden rows at the top and bottom are where the real money is, and they are the two nobody invoices you for. Everything in the middle shows up in a bank statement eventually; the first and last rows show up only in a funnel that quietly underperforms. The same two rows are the reason the steady-state figure in what an in-house sales and retention team really costs understates a high-churn desk.

Ramp: the weeks before a new hire goes near a live client

Write your ramp down as a sequence with a named owner and a duration for each step. Until it is written, "two weeks training" is a story the desk tells itself.

1. Paperwork and access. Contract, background checks, CRM, telephony and recording logins.
2. Product and platform. What you offer, how accounts are funded, what the fees and charges are.
3. The jurisdiction map. Who this desk may lawfully contact, on whose licence, and who it may not.
4. Approved wording and disclosures. The script, the required risk warnings, and what may never be said.
5. Onboarding mechanics. Verification steps, document handling and the data rules that apply to them.
6. Supervised calling. Listening first, then calling with a reviewer able to stop the call.
7. Sign-off. A named supervisor records that this person may work live clients unsupervised.

Two numbers come out of this and both belong in the cost stack. The first is ramp duration: calendar days from day one to sign-off. The second is ramp cost: full pay and employer cost across those days, plus the reviewer and supervisor hours consumed. A desk that cannot say what its ramp duration is cannot price its attrition, and usually discovers the answer is longer than the training plan claims.

One discipline is worth defending against every quarter-end. Step seven is not a formality. Letting an unsigned-off hire work live clients because the desk is short is how script drift starts, and script drift in this industry is a conduct issue, not a quality issue. If pressure to skip it is recurring, the problem is capacity, and a leaner sales team structure is the better conversation.

The leads and clients that die in the handover

The most expensive consequence of a departure is rarely on the HR ledger. It is the book: the enquiries this person had in progress, the ones they had not reached yet, and the clients who were used to speaking to them. During the notice period, the gap and the ramp, that book is worked badly or not at all.

Speed matters more than the industry generally admits. The work by James Oldroyd with co-authors, published in Harvard Business Review in March 2011 as "The Short Life of Online Sales Leads" and related to the earlier MIT and InsideSales lead response research, found that contacting web-form leads within minutes rather than hours sharply improved the odds of reaching and qualifying them; the bibliographic record is held at BYU ScholarsArchive. That data is American, from the 2007 to 2011 period, drawn from web-form business leads and not from financial services in India or the Gulf, so take the direction and not the multipliers: a lead that waits does worse, and a lead that waits through a vacancy was paid for twice over. The practical consequences of that are in speed to lead for forex brokers.

Three handover rules cost nothing and save most of this. Reassign the whole book the day notice is given, not the day the person leaves. Have the departing person make a short, recorded handover call on their live relationships rather than sending a spreadsheet. And flag every record untouched for longer than your own follow-up standard, so the orphans are visible in a report instead of discovered three months later. If your CRM cannot do those three things, that is a tooling gap worth fixing before the next resignation.

A worked example with round placeholders

Illustrative, replace with your numbers. Take a desk of fifteen phone seats. Assume gross pay of Rs 20,000 a month per seat as a round placeholder, and assume a ramp of six weeks from day one to sign-off.

• Ramp pay: roughly 1.5 months of pay and employer cost before the seat is fully productive
• Recruiting: your own recruiter fees and job-board spend from last year, divided by the number of hires it produced
• Interview, onboarding and supervisor hours: hours x the cost per hour of the people who spent them
• Orphaned book: the number of records unworked during the gap, valued at what you paid to acquire them
• Annual bill: cost per replacement x the number of seats that turned over last year

The formula to keep is: annual churn cost = (notice-period loss + recruiting + selection hours + onboarding + ramp pay + supervisor hours + orphaned book) x seats replaced in the year. Run it once with your real numbers and the result usually reframes the argument about incentive budgets and lead supply, because both are typically smaller than the churn bill they could prevent.

Then run one sensitivity. Hold everything constant and cut ramp duration by a third, which is usually achievable by writing the ramp down and resourcing the reviewer properly. Then hold ramp constant and cut replacements by a third. Whichever lever moves the number more is the one to work on first, and for most desks that have never priced this, it is the second.

Indore and tier-2 desks: cheaper to hire, longer to ramp

Portal data puts caller and sales pay in Indore and comparable tier-2 Indian cities below metro levels, which lowers two lines in the stack directly: ramp pay and the per-seat cost of the hours you lose. That is a genuine advantage and it is why many brokers and IBs run back offices there rather than in a metro.

The offsetting effects are equally real. You will mostly recruit from general telecalling, insurance and retail sales backgrounds rather than from people who have worked a regulated financial desk, so steps three, four and five of the ramp above take longer and need better material. The local market for experienced staff is thinner, which means a departure is harder to replace quickly and more likely to be filled with somebody who needs a full ramp rather than a short one. And the supervision load does not fall with the rent: a satellite desk needs more structured listening and review, not less, because the people who set the standard are not sitting in the room.

There is a compliance dimension that belongs here rather than in a legal appendix. A desk that turns over quickly, hires from pools without financial-services experience and is supervised at a distance is exactly the configuration in which approved wording drifts into the promises regulators have publicly warned about. Treat review hours in a tier-2 back office as a fixed cost of operating it, not one that scales down with the salary bill.

The same logic travels. A UAE desk sits in a higher-cost labour market and its ramp content differs by regime, since onshore, DIFC and ADGM each carry their own promotion and client-classification requirements to teach. The structure of the stack does not change; only the size of the lines does.

Cutting the churn bill without pretending nobody leaves

Some turnover is healthy and none of it is avoidable. The goal is a smaller bill per departure and fewer departures caused by things you control.

Four moves do most of the work. Write the ramp down and resource the reviewer, which shortens the most expensive line. Stabilise incentive pay so that a weak month in lead supply does not read as a pay cut to the caller, which removes one of the common push factors. Fix the handover rules above, which protects the book. And address lead flow, because a desk grinding through poor-quality cold outreach loses people faster than one working enquiries with real intent.

That last point is where sourcing meets retention. OptimizedLeads supplies ready leads on a plan through its forex leads service, worked from existing data rather than a fresh campaign each month, with a free CRM to track, assign and follow up so that a departing person's book is reassignable in minutes rather than reconstructed from a spreadsheet, and invalid leads replaced on terms agreed in writing. Confirm the definition of an invalid lead and the replacement window in writing before signing. Plans are at /pricing and the two-day trial at /start-free-trial will tell you more about fit than any proposal will.

Frame the restructuring honestly. The aim is to redeploy people onto the work that actually wins a funded client and to stop adding seats as targets rise, not to cut the team. Scaling without scaling headcount covers the growth side, and the transition plan from in-house calling to a lead provider covers the sequencing if you decide to change the model. If your desk is small, stable and well utilised, keep it; when in-house calling stops making sense describes the point at which that stops being true.

What to do this week

• Build an exit register for the last twelve months: seat, role, tenure, reason given, days vacant.
• Write your ramp down as steps with an owner and a duration, then measure the real elapsed time of the last hire.
• Price one replacement end to end using the stack above, with your own invoices and hourly costs.
• Multiply by last year's replacements to get the annual churn bill, and put it next to your incentive budget.
• Count the records that sat unworked during your last vacancy, and value them at what you paid for them.
• Check that no current hire is working live clients without a recorded sign-off.

This is operational and cost-accounting guidance, not legal, HR or investment advice, and nothing here is a recommendation about trading. Employment rules, statutory rates and regulator positions change: verify with a qualified lawyer, your CA and the relevant regulator before acting.

Sources and further reading

• Aon — 32nd annual salary increase and turnover survey, India attrition and 2026 increase projection.
• SHRM — United States cost-per-hire reporting and a consultant's replacement-cost estimate, used as structure only.
• BYU ScholarsArchive — bibliographic record for "The Short Life of Online Sales Leads", Harvard Business Review, March 2011.
• Reserve Bank of India — press release of 3 February 2022 on authorised persons, authorised platforms and agents promising exorbitant returns.

Frequently Asked Questions

What does it cost a broker to replace one sales or retention seat?

Price it end to end rather than looking for a benchmark. Add notice-period output loss, recruiter fees and job ads, interview and selection hours, onboarding and access setup, ramp pay until sign-off, supervisor and reviewer time, and the value of the leads and clients left unworked. Multiply by the seats you replaced last year for the annual bill.

Is there a reliable attrition figure for forex sales teams in India?

No credible, independently sourced figure for forex sales or retention roles was found. Aon's cross-industry survey put overall Indian attrition at 16.2% in 2025, which is useful context but not a planning assumption for a broker desk. Build the figure from your own exit register.

How long should ramp take before a new hire works live clients?

That depends on your product and your rulebook, so measure rather than assume. Write the ramp as steps with owners and durations, covering access, product, the jurisdiction map, approved wording and disclosures, onboarding mechanics, supervised calling and a recorded sign-off, then time the last hire against it.

Why does attrition hurt more on a regulated desk?

Because a replacement cannot simply be handed a phone. They must be trained on who may lawfully be contacted and on whose licence, on approved wording and required disclosures, and on onboarding and verification steps. Short-cutting that under pressure is how script drift and conduct problems begin.

How do I protect the leads of someone who resigns?

Reassign the whole book on the day notice is given rather than on the last day, have the departing person make recorded handover calls on live relationships, and run a report that flags any record untouched for longer than your follow-up standard. A CRM that supports reassignment in minutes removes most of the loss.

Want to see whether better lead flow lowers your churn bill? Start the free 2-day trial or read how the forex leads service works.