Inside the romance scam economy: how pig butchering rings actually work

When most people picture a romance scammer, they picture an individual: a person at a laptop somewhere, running one or two cons. The reality, for the operations that produce the bulk of romance-fraud losses globally, is closer to a call centre than a lone con artist. There are buildings. There are managers. There are training manuals. There are KPIs.

This post draws on reporting from journalists, NGOs, and law-enforcement agencies who have, over the past three years, mapped the structure of pig-butchering operations across Southeast Asia. Most of what follows is now public knowledge; some of it will be uncomfortable.

The compounds

The largest concentrations of pig-butchering operations are in special economic zones in Cambodia, Myanmar, and Laos — particularly in border regions with weak law enforcement and poor extradition treaties. The infrastructure is industrial: multi-storey buildings, sometimes converted hotels or casinos, housing hundreds to low thousands of workers each. Sihanoukville in Cambodia and the Shwe Kokko complex in Myanmar are the most-reported examples. There are dozens of smaller sites.

Many of the workers are themselves victims of human trafficking. They are recruited under false pretences — usually offered legitimate-sounding tech or customer-service jobs in Thailand, Cambodia, or Singapore — then have their passports confiscated on arrival and are forced to run scams against quotas. Workers who fail to meet quotas are reported to face physical abuse. Workers who try to leave are sometimes resold to other compounds.

This is not a metaphor. This is what the United Nations Office on Drugs and Crime documented in its 2023 and 2024 reports. The romance scam you may have personally encountered was very likely run by someone who is themselves a captive worker.

The operational structure

A typical operation has four roles:

Recruiters — handle phase one (the cold approach). They run high volumes of "wrong number" texts and dating-app matches simultaneously, qualifying targets and warming them to the point where they can be passed to a relationship operator.

Relationship operators — handle phase two (the weeks-long rapport). Each operator runs 5-15 active "relationships" simultaneously. They work from scripts but personalise. They are often the captive workers; quality of English correlates with how senior or how trafficked they are.

Closers — handle phase three and four (the demonstration and the slaughter). These are usually the most experienced operators, and they take over a "ripe" target from the relationship operator for the actual money extraction. They are paid on commission.

Tech and laundering — build and maintain the fake investment platforms, manage the crypto wallets, and move extracted funds through tumbling services and into the legitimate financial system. This is where most of the senior money is made.

The economics: a single operation can extract tens of millions of dollars per year. The organisations behind them are believed to be primarily Chinese organised-crime syndicates with regional partnerships. The combined annual revenue across the industry is estimated, by the US Institute of Peace, at over $75 billion globally in 2024. That is a larger illicit revenue stream than the global cocaine trade.

Why scripts matter

If you've ever felt that something a scammer said sounded canned — a particular phrase about destiny, a specific way of describing their dead spouse, an oddly precise timeline of when they want to meet you — you weren't imagining it. The scripts are written, tested, and iterated on like marketing copy. The phrases that produce the best engagement are kept and circulated; the ones that don't are dropped.

This is also why patterns repeat across thousands of victims who never spoke to each other. The "I'm in finance, my uncle taught me to trade crypto, let me show you how it works" arc is not coincidence. It is a script that has been A/B tested across hundreds of thousands of conversations and optimised.

The economics of detection

A reasonable question: if these operations are this large and this well-known, why aren't they shut down?

The honest answer has three parts. First, jurisdictional: the compounds operate in places where local enforcement is corrupt, captured, or unwilling to act. The criminal organisations have integrated with regional political structures. Cracking down requires multilateral cooperation that has, so far, been slow.

Second, the money flows are sophisticated. By the time funds leave the victim's bank account, they have moved through multiple shell companies, crypto exchanges, and tumbling services within hours. Recovery is rare even when the perpetrators are identified.

Third — and this is the uncomfortable part — the volume of victims is so high, and individual losses so distributed, that the political pressure has historically been lower than for, say, ransomware attacks on hospitals. This is changing as total losses climb past $75bn globally and as victims include, increasingly, professionals with media access. But the response is still lagging.

What a defensive layer looks like

If you accept that the supply side of romance fraud is industrial, sophisticated, and not going away in the medium term, the question becomes: what does an effective demand-side defence look like?

The honest answer is: education and friction. Public awareness of pig-butchering structure has roughly doubled in the past two years, and reported losses-per-victim in the UK have started to plateau as a result. Tools like ours add friction at the verification step — the moment a user could check the photos, the story, the platform — and that friction matters at scale, even when individual cases still slip through.

The thing that doesn't work is moralising at victims. The scams are engineered to bypass the cognitive defences of competent adults. Telling a victim "you should have known better" is both factually wrong (the operation was specifically designed to defeat what they knew) and counterproductive (it pushes the next victim into not telling anyone, which gives the scammer more time).

Where TruthHound fits

We are part of the friction layer. We can scan a profile, a chat, a voice note, and tell a user — in three minutes — what we see. That changes the cost-benefit of due diligence from "an hour of awkward Googling" to "a few minutes and a clear report you can show a friend." That is a small wedge against a very large industry. It is also the only kind of wedge that operates at the speed and scale needed.

We are not solving romance fraud. The compounds will not close because of TruthHound. But the conversation a parent can have with a child, or a friend can have with a friend, when there is a written, evidence-backed report on the table — that conversation is more likely to end with the person walking away from the scam than the same conversation without the report.

That, modestly, is the thing we built.