kyccost

Independent reference. Not legal or regulatory advice. Consult a qualified compliance specialist for advice specific to your jurisdiction and risk profile. See methodology.

Cluster 6 / Hidden labour

The hidden labour cost of KYC alerts.

False-positive content is dominated by AML transaction-monitoring vendors. The onboarding-stage labour cost (sanctions hits at signup, PEP matches, document anomaly flags) is a discrete cost line that rarely surfaces. This page sizes it.

False-positive rates: 80-95% legacy | Investigation: 6-25 min / £2.20-£11.70 per alert | Per-customer ops labour: £6-£22

The 95% problem, applied to KYC.

Lucinity, LSEG and ACAMS routinely cite false-positive rates of 80-95% on legacy AML transaction-monitoring systems. The same effect applies at onboarding stage. Sanctions screening, PEP screening and adverse media screening generate alerts that are almost entirely noise; the labour cost of investigating those alerts is the line vendor pricing pages cannot quote because vendors do not bear it.

The onboarding-stage figure is sometimes treated as a smaller version of the AML monitoring problem. It is not. Onboarding alerts must be cleared inside the customer's acceptable wait time; an alert that takes 24 hours to resolve at AML monitoring is acceptable, the same alert at onboarding causes drop-off and CAC inflation. The product hit-rate × per-alert-cost is consequently the dominant line on per-customer KYC for any fintech with a marketing-led growth motion.

What an onboarding alert actually costs.

Alert typeAvg minutesHourly rate (UK)Per-alert cost
Sanctions hit (junior triage)8£24£3.20
PEP match (junior triage)10£24£4.00
Adverse media flag (junior triage)15£24£6.00
Document anomaly (junior + tooling check)12£24£4.80
Senior MLRO escalation (true-hit case)60£130£130.00
Source-of-funds review (EDD case)75£40£50.00

UK fully-loaded labour rates per ONS data for compliance occupations, averaged across major fintech employers (Q1 2026). Senior MLRO rate is mid-band; chartered MLRO with PEP / sanctions specialism runs higher.

Worked example: 50,000 onboardings.

Assumptions
  • 50,000 annual onboardings
  • 4% combined sanctions / PEP / adverse media hit rate
  • £6.50 average junior investigation cost
  • 5% of hits escalate to senior MLRO at £130 / hour
  • 1% of cases trigger SoF review at £50 / case
Build-up
Junior alert investigation (50,000 × 4% × £6.50)£13,000
Senior MLRO escalation (5% × 2,000 hits × £130)£13,000
SoF review (1% × 50,000 × £50)£25,000
Total ops labour, year 1£51,000
Per onboarded customer£1.02
Excludes EDD overlay labour, ongoing monitoring labour, periodic refresh, and QA. The full per-customer labour line typically lands £6-£22 once those are added.

Drop-off cost is a real cost.

False positives produce friction in the onboarding flow. A customer routed to a manual review queue waits hours or days; the marketing-led acquisition motion does not survive that wait at scale. Industry benchmarks suggest onboarding drop-off inflates fintech CAC by 30-50% beyond marketing-only benchmarks (Prospeo 2026 fintech CAC dataset; Sumsub Drop-off Index commentary).

A 1-percentage-point reduction in false-positive rate, on a 100,000-onboardings book, recovers roughly 1,000 customers worth of CAC at marketing-only rates. At fintech LTV multiples, the recovered LTV often exceeds the labour saving by 3-5x. The downstream economics of false-positive reduction are consequently larger than the labour-line analysis suggests.

Where automation reduces it.

Lucinity, LSEG and ComplyAdvantage publish case studies citing 60-70% false-positive reductions with AI-assisted triage on AML monitoring and sanctions screening. The savings are real but offset by model risk-management cost: validation, monitoring, governance, model documentation. Net saving typically 35-55% of the original labour line at scale, plus the engineering / data-science build cost amortised over the contract period.

For fintechs onboarding above 100,000 customers a year, AI-assisted triage on adverse media screening alone usually pays back within 12-18 months. Below that volume, the build cost typically dominates the saving. The build-vs-buy decision applies to the triage layer specifically and is usually distinct from the platform-level decision; see build vs buy.

FTE sizing calculator.

Onboarding volume × hit rate × minutes per alert returns a defensible alert-investigation FTE count. Add an estimated 80-150% on top for the rest of the ops team (EDD case work, ongoing monitoring, escalation, periodic refresh, QA).

FTE sizing inputs
100,000
4%
8 min
£24
Output
Total annual alerts
4,000
Total annual labour hours
533
Required FTE (alert review only)
0.3
at 1,700 productive hours per FTE per year
Annual labour cost
£12,800
Excludes senior MLRO escalation, EDD case work, periodic refresh and QA. Add 80-150% for fully-loaded ops team.

False-positive cost questions

How much do false positives cost in KYC?+
At onboarding stage, false-positive labour typically costs £6-£22 per onboarded customer at industry-typical hit rates. The product is hit-rate × investigation-cost: 4% sanctions / PEP / adverse media hit rate × £6.50 average investigation cost = £0.26 of pure labour per onboarding, plus senior MLRO escalation on the 3-7% of cases that need it. For a 100,000-customer book the annual labour line typically lands £200,000-£600,000.
What percentage of KYC alerts are false positives?+
Industry benchmarks (Lucinity, LSEG, ACAMS, ComplyAdvantage) place sanctions and PEP false-positive rates at 80-95% on legacy keyword-matching systems. Adverse media false-positive rates run higher still, typically above 90%. AI-assisted triage reduces false positives 60-70% in published case studies; the saving is real but offset by model risk-management cost and the residual review labour on the cases the model surfaces.
How much labour does AML alert review require?+
Typical 6-25 minutes per junior-analyst alert investigation; 30-90 minutes per senior MLRO escalation. UK fully-loaded labour rates: £18-£28/hour junior analyst, £85-£180/hour senior MLRO. Per-alert cost typically £2.20-£11.70 junior, £40-£270 senior. The alert mix matters more than the per-alert cost: a programme that escalates 5% of cases to senior MLRO costs materially more than one that escalates 1%.
How does false-positive rate affect onboarding cost?+
False positives directly drive ops labour cost (the dominant line on per-customer KYC) and indirectly drive customer drop-off cost (false positives produce friction in the onboarding flow). Industry benchmarks suggest onboarding drop-off inflates fintech CAC by 30-50% beyond marketing-only benchmarks. The combined effect is that a 1-percentage-point reduction in false-positive rate is worth more in fintech P&L terms than most automation investments would suggest.
Can automation actually reduce false-positive rates?+
Yes, with caveats. Lucinity, LSEG and ComplyAdvantage published case studies cite 60-70% false-positive reductions with AI-assisted triage. The caveats: model risk-management cost (validation, monitoring, governance), data labelling cost on the training set, and the residual review labour on the cases the model surfaces are real offsetting lines. Net saving is typically 35-55% of the original labour line at scale.
How many KYC analysts does a fintech actually need?+
For a 100,000-onboardings-per-year book at 4% hit rate and 8 minutes per alert, the total alert-investigation labour is roughly 530 hours per year, which is one-third of an FTE. Adding ongoing monitoring, EDD case work, escalations, periodic refresh and quality assurance lifts the figure to 3-7 FTE for a mid-sized EMI; 8-18 FTE for a 250,000-onboardings retail fintech; materially higher for crypto exchanges and brokers.

Sources cited on this page

  1. Lucinity Real Cost of AML Compliance, 2025 commentary
  2. LSEG / Forrester True Cost of AML Compliance, most recent edition
  3. ACAMS commentary on false-positive rates in sanctions and PEP screening
  4. ComplyAdvantage published reduction case studies
  5. ONS UK compliance occupation labour-rate benchmarks (2025-2026)
  6. Sumsub Drop-off Index commentary