Case study · Trust & payments · Freelance platform · 2021–2023
Trust and payouts: KYC, rule-based fraud triggers and payout redundancy cut fraud from hundreds of cases a month to under 10
Head of Freelance Platform · fraud rules with the CTO · 2021–2023
- < 10fraud cases a month, from hundreds
- 70%of freelancers verified on 2+ payout methods
- 4specialist vendors integrated end to end
Situation
Freelancers in 80+ countries took client work and withdrew earnings through the platform. Two things had to be true: clients had to trust who was doing the work, and freelancers had to trust they’d be paid.
Complication
Fraud ran at hundreds of cases a month, identity checks were not built for sensitive data at scale, and payouts relied on a narrow set of providers, so one provider’s policy change could cut off pay for a whole region. Sanctions rules added a compliance duty that needed its own controls.
Decision
Buy the commodity, build the edge. Identity checks, e-signing, originality checks and training are done better, and regulated more strictly, by specialists, so we bought them. Fraud rules, risk profiles and payout logic were built in-house, because that is where the marketplace competes and where its own data is the advantage.
Exhibit
We bought what specialists do better and built what makes the marketplace different
↑ Specialist or regulatory complexity→ Differentiation for the marketplace
High specialist or regulatory complexity · low differentiation for the marketplace
- Veriff (KYC)
- DocuSign (e-signing)
- CopyLeaks (originality)
High specialist or regulatory complexity · high differentiation for the marketplace
- Fraud rules and risk profiles
- Payout logic and verification
Low specialist or regulatory complexity · low differentiation for the marketplace
- Teachable (training)
Low specialist or regulatory complexity · high differentiation for the marketplace
- Pricing and matching
- Buy: Veriff (KYC) (low x, high y)
- Buy: DocuSign (e-signing) (low x, high y)
- Buy: CopyLeaks (originality) (low x, mid y)
- Buy: Teachable (training) (low x, low y)
- Build: Fraud rules and risk profiles (high x, high y)
- Build: Payout logic and verification (high x, mid y)
- Build: Pricing and matching (high x, low y)
What I did
- 01
Verified identity. Designed the platform-security and KYC system: Veriff verification, freelancer risk profiles, 2FA on withdrawals and document retention for sensitive identity data.
- 02
Turned fraud into rules. With the CTO, rule-based fraud triggers, fraud categories and a response playbook for each category.
- 03
Removed single points of failure in payouts. Launched Wise and wire transfers, ran a multi-source Payoneer setup, retired PayPal and Skrill, and added payment-details verification.
- 04
Built sanctions compliance into the funnel. Restrictions and blocking for sanctioned countries, a document-retention policy and timestamped re-signing of terms.
- 05
Ran every integration end to end. From vendor evaluation and build-vs-buy to API implementation with engineering: Veriff, DocuSign, CopyLeaks, Teachable.
Impact
- Fraud from hundreds of cases a month to under 10.
- 70% of freelancers verified on 2+ payout methods, so no single provider can stop supply.
- KYC, 2FA and sanctions controls on every withdrawal.
Learning
In a marketplace, payout reliability is a supply metric.